Del Fiol G, et al. J Am Med Inform Assoc 2015;22:519–528. doi:10.1136/amiajnl-2014-003127, Research and applications
Data standards to support health information exchange between poison control centers and emergency departments
RECEIVED 9 July 2014 REVISED 3 October 2014 ACCEPTED 12 October 2014 PUBLISHED ONLINE FIRST 23 October 2014
Guilherme Del Fiol1, Barbara Insley Crouch2, Mollie R Cummins1,3
ABSTRACT ....................................................................................................................................................
.................................................................................................................................................... Key words: poison control, clinical decision-making, health information exchange, electronic health records
INTRODUCTION
routinely call poison centers for consultation. In a typical patient consultation, both PCC specialists and ED providers place several phone calls throughout the ED admission. In these phone calls, a significant amount of information is exchanged such as a case summary, description of the exposure, the patient’s current status, diagnostic test results, and recommendations for treatment. In previous work, we identified a number of inefficiencies and safety vulnerabilities in the telephone-based ED–PCC communication process.5 These inefficiencies and safety vulnerabilities included repetition of redundant clinical information, vague or ambiguous communication of clinical observations and diagnostic test results, and difficulty establishing telephone communication due to provider unavailability. Through a nationwide Delphi study, we also identified widespread support for a health information exchange (HIE)-supported communication process.6 Based upon the results of that study, we propose a new process in which HIE does not entirely eliminate
Poisoning deaths have increased dramatically over the last 10 years, surpassing firearm deaths in 2004 and motor vehicle traffic deaths in 2011. The dramatic increase in deaths is largely due to medication overdose, especially opioid pain medications.1 Emergency department (ED) visits for poisoning have also dramatically increased during the same period.2–4 US poison control centers (PCCs) play a critical role in the management of poison exposures. These centers are staffed 24 h a day by specialists in poison information who are health professionals specially trained in toxicology. These specialists assess poison exposures over the telephone and provide treatment recommendations. This service is available to both healthcare providers and the general public. PCCs also collect detailed information about exposure cases in order to provide public health services and poison surveillance. Given the sheer diversity of potential poisons and the complexity of managing unusual and multi-substance exposures, ED care providers
Correspondence to Dr Guilherme Del Fiol, Department of Biomedical Informatics, University of Utah, 421 Wakara Way, Suite 140, Salt Lake City, UT 84108, USA; E-mail: [email protected] C The Author 2014. Published by Oxford University Press on behalf of the American Medical Informatics Association. V All rights reserved. For Permissions, please email: [email protected] For numbered affiliations see end of article.
519
RESEARCH AND APPLICATIONS
Objective Poison control centers (PCCs) routinely collaborate with emergency departments (EDs) to provide care for poison-exposed patients. During this process, a significant amount of information is exchanged between EDs and PCCs via telephone, leading to important inefficiencies and safety vulnerabilities. In the present work, we identified and assessed a set of data standards to enable a standards-based health information exchange process between EDs and PCCs. Materials and methods Based on a reference model for PCC–ED health information exchange, we (1) mapped PCC–ED information exchange events to clinical documents specified in the Health Level Seven (HL7) Consolidated Clinical Document Architecture (C-CDA) Standard, and (2) mapped information types routinely exchanged in PCC–ED telephone conversations to C-CDA sections. Results Four C-CDA document types were necessary to support the PCC–ED information exchange process: History & Physical Note, Consultation Note, Progress Note, and Discharge Summary. Information types that are commonly exchanged between PCCs and EDs can be reasonably well represented within these C-CDA documents. Conclusions A standards-based health information exchange process between PCCs and EDs appears to be feasible given a set of clinical data standards that are required for EHR certification in the USA, although the proposed approach still needs to be validated in actual system implementations. Such a process has the potential to improve the safety and efficiency of PCC–ED communication, ultimately resulting in improved patient care outcomes.
Del Fiol G, et al. J Am Med Inform Assoc 2015;22:519–528. doi:10.1136/amiajnl-2014-003127, Research and applications
standardized terminologies.10,11 Consequently, NPDS data are valuable for poisoning surveillance but are poorly interoperable from a clinical informatics perspective.
RESEARCH AND APPLICATIONS
telephone consultations, but supports and reduces the need for synchronous telephone communication. HIE is potentially valuable in supporting continuity of care as patients transition from one healthcare setting to another, especially when transitions of care occur across healthcare organizations without interoperable information systems. Currently, we are developing a scalable, standards-based approach to HIE between EDs and PCCs as part of an AHRQ (US Agency for Healthcare Research and Quality)-funded study to evaluate the effect of HIE-supported communication on workflow, communication, efficiency, and utilization.7 This type of HIE is distinctive in that it will involve bidirectional information sharing and communication in support of an emergency healthcare event, as it unfolds. This will entail multiple HIE messages that support real-time collaboration and decisionmaking. The goal of the present work was to identify and assess a set of data standards to support dynamic electronic information sharing between EDs and PCCs. We focused on standards that are required for EHR Meaningful Use certification in the USA, specifically the Health Level Seven (HL7) Consolidated Clinical Document Architecture (C-CDA) Standard.8
PCC–ED information types In previous work, we analyzed call recordings corresponding to a random sample of 120 poison exposure cases entailing telephone communication between an ED and the Utah Poison Control Center.5,12 Using qualitative methodology, we identified 52 unique information types exchanged during ED–PCC telephone communication. A subset of these information types that we selected for mapping to the C-CDA standard are described in table 1. These information types, derived from real-world communication between EDs and PCCs and well-validated, encapsulate the focused subset of information currently used to support collaboration.12 Several of the information types closely correspond to NPDS data elements, since communication partly serves to support collection of information for entry into the poison center information system, which collects NPDS data elements. However, the information types used for mapping reflect the overall ED–PCC communication that currently takes place in support of both patient care and population health.
BACKGROUND AND SIGNIFICANCE Collaboration and communication between PCCs and EDs Collaboration between an ED and a PCC occurs in three phases, described in detail in Cummins et al,5 and can be initiated by either party. In some cases, the PCC refers a patient to an ED for further evaluation and treatment. In other cases, a patient presents to an ED without contacting the PCC, and collaboration is initiated by an ED physician, nurse, or other staff member. Over the telephone, the PCC specialist assesses the poison exposure, provides consultation, and recommends a plan of care. Communication is ongoing as the case unfolds. The PCC calls periodically to request information on the patient’s clinical condition, diagnostic test results, treatment, and plan of care, updating treatment recommendations as necessary. The purpose of ongoing telephone communication is twofold: (1) to support ongoing emergency treatment for poison exposure; and (2) to support the collection of case data for both local (state and local health departments) and national poisoning surveillance (National Poison Data System, NPDS).9 However, reliance upon telephone communication leads to inefficiencies and safety vulnerabilities.5
Consolidated Clinical Document Architecture Standard The HL7 CDA standard provides a framework for representing various kinds of clinical documents, such as progress notes, discharge summaries, and imaging reports.13 CDA documents consist of a header and a body. The header includes document metadata, such as document type, patient demographics, care provider and/or healthcare institution responsible for generating the document, and document timestamp. The body contains the patient’s clinical data, which are organized into multiple sections. Each section is modeled based on standard reusable templates, such as plan of care, problems, medications, and results. Therefore, the CDA body is constructed with modular blocks of data, with standard templates providing the building blocks. The C-CDA standard is an HL7 implementation guide that builds upon CDA release 2.0. It defines nine standard document types for different clinical use cases: Continuity of Care Document, Consultation Note, Discharge Summary, Procedure Note, Diagnostic Imaging Reports, History and Physical, Operative Note, Progress Note, and Unstructured Documents.8 The C-CDA standard is required for EHR certification in the USA EHR Meaningful Use program.14 Since the C-CDA standard was designed to cover a broad range of clinical use cases, it is a reasonable choice to support PCC–ED HIE.
Data elements for poison exposure information Using a proprietary, purpose-built information system, the PCC collects structured and narrative data describing each poison exposure case. The structured data largely consist of NPDS data elements. The NPDS data elements allow for the collection of highly detailed information, in both structured and narrative form, about the nature and circumstances of each poison exposure. The NPDS data elements are not represented in any standard information model and coded data are not represented in standard controlled terminologies. In previous work, we identified that these data elements only partially map to
MATERIALS AND METHODS Our approach entailed three steps. First, we designed a reference model for PCC–ED HIE events. This model was informed by our previous work, in which we systematically observed and analyzed PCC–ED phone conversations.5 Second, we mapped each PCC–ED HIE event to a set of candidate C-CDA document types. This mapping was informed by the document definitions
520
Duration of exposure (acute vs chronic) Narrative information about the poison exposure scenario specifically relevant to determining the certainty of ingestion/exposure, and general information helpful in constructing a clinical picture of the patient and treatment plan General grouping or characterization of a poison (eg, b-blocker) General characteristics of a poison (eg, half-life, mechanism of action, peak effect) Generic name of a poison that is also a therapeutic agent Dosage or amount of substance (eg, 500 mg metformin). Brand name of a poison that is also a therapeutic agent Informal description ranging from general drug class (eg, antiarrhythmic) to intended purpose (eg, a chemical used to clean carburetors) For example, tablet, pill, powder, lozenges For example, extended release vs rapid release Narrative describing reasoning of a PCC specialist in identifying a substance based on characteristics of the poison (eg, ‘A blue pill in that shape can be a few different things, but it is probably Viagra’) Name of a non-pharmacological substance For example, solid, inhalant, liquid Amount of time elapsed since the initial poison exposure (eg, ‘It has been about 5 h since she took the pills’)
Chronicity
Establishing background/certainty
Substance class
Substance information
Substance name (generic)
Substance amount
Substance name (brand)
Substance description
Substance form
Substance formulation
Substance identification rationale
Substance–non-pharmacological
Substance type
Time since ingestion
Information about the patient’s past medical history Medications that the patient currently takes at home
Medical history
Patient medications
(continued)
The level of certainty about the precise formulation, based on the subjective/objective information provided in a circumstance where the poison is a therapeutic agent
Certainty of formulation
RESEARCH AND APPLICATIONS
521
Patient health history
For example, polypharmacy, accidental exposure, overdose
Exposure type
Exposure information
Table 1: Information types identified in an observational study of ED–PCC telephone communication* Del Fiol G, et al. J Am Med Inform Assoc 2015;22:519–528. doi:10.1136/amiajnl-2014-003127, Research and applications
Signs observed by ED healthcare providers Healthcare provider or PCC assessment that a symptom is unrelated or likely unrelated to the poison exposure. For example, a patient may exhibit tremors related to underlying Parkinson’s disease, which are unrelated to an acute narcotics overdose Information describing the patient’s blood pressure, heart rate, and/or respiratory rate along with the date and time the vital signs were obtained
Physical exam findings
Unrelated symptoms
Vital signs
522
Specific dose or amount of exposure at which a substance becomes toxic Circulating blood level of a substance considered toxic Potential or expected clinical effects of a given poison exposure Description of the most harmful clinical effects and poorest outcome that a patient might experience (a description of risk)
Toxic dose
Toxicity levels
Clinical effects of substance
Worst case scenario
Medications prescribed for patient at time of discharge from the ED ED description of the patient’s clinical condition, particularly whether clinical effects of exposure are observed ED healthcare provider plan of care for the patient (eg, treatments, procedures, length of stay, parameters for discharge) Inquiry about results of diagnostic tests (eg, laboratory, ECG, imaging) Information about the ED’s planned or recommended timing of subsequent diagnostic testing (eg, ‘We’ll get another level at 4 h [after ingestion]’) Information about the timing of a treatment or a laboratory test, usually in relation to the time of ingestion (eg, ‘That level was drawn 2 h after ingestion’) Information that a treatment, directly related to the poison exposure, was administered
Patient discharge medications
Patient status
Plan of care
Diagnostic test results
Time next laboratory tests will be ordered
Time laboratory test was performed/drawn
Treatment performed
*Adapted from Cummins et al. One goal of telephone communication between EDs and PCCs is the collection of information for entry into the NPDS. Consequently, some information types resemble NPDS data elements. However, NPDS data elements are defined differently, according to a published coding manual.18 ED, emergency department; NPDS, National Poison Data System; PCC, poison control center.
12
ED staff member indicates whether or not a PCC recommended treatment has been administered to a patient
Confirmation that treatment was given
ED treatment/management information
PCC recommendations for treatment and/or duration of direct observation prior to discharge from the ED
PCC recommendations for treatment and discharge parameters
PCC recommendations and toxicology information
Information about the patient’s mental status Symptoms as reported by the patient or caller to the PCC
Caller reported symptoms
The absence of any signs or symptoms attributable to the poison exposure
Absence of clinical effects
Mental status
Patient’s chief complaint upon presenting to the emergency department or calling the PCC
Chief complaint/reason for visit
Subjective and objective information
RESEARCH AND APPLICATIONS
Table 1: Continued
Del Fiol G, et al. J Am Med Inform Assoc 2015;22:519–528. doi:10.1136/amiajnl-2014-003127, Research and applications
Del Fiol G, et al. J Am Med Inform Assoc 2015;22:519–528. doi:10.1136/amiajnl-2014-003127, Research and applications
provided in the C-CDA specification. Next, we mapped each PCC–ED information type to C-CDA sections contained in the documents identified in the previous step. This mapping was informed by the section and template definitions and examples provided in the C-CDA specification. We excluded information types that constituted questions or verbal requests for information, specific to telephone-based communication (eg, confirmation that the patient has arrived at the ED, inquiry regarding the patient’s discharge or transfer status), and information types routinely included in C-CDA headers. Last, for each PCC–ED HIE event, we selected the C-CDA document whose sections best matched the PCC–ED information types. Two investigators with expertise in HL7 standards (GDF) and PCC–ED information exchange (MRC) independently completed mappings. We reconciled the independent mappings iteratively and through consensus with a poison control expert (BIC).
different ways: (A) The PCC receives a phone call from a patient and refers that patient to the ED by sending an initial C-CDA Consultation Note; or (B) the ED receives a poison-exposed patient and requests consultation from the PCC through a C-CDA History & Physical Note. Once a case consultation is established, a set of iterative events take place depending on the patient’s progress: the ED submits periodic patient status updates to the PCC using the Progress Note C-CDA (C); and the PCC submits any updated management recommendations to the ED using the Consultation Note C-CDA (D). Finally, when the patient is transferred or discharged, the ED sends a C-CDA Discharge Summary document to the PCC, closing the case (E). Table 2 describes each of the PCC–ED HIE events.
RESULTS Reference model for PCC–ED HIE events with mappings to C-CDA document types Figure 1 depicts the HIE events necessary to support ED–PCC communication and collaboration. The process may begin in two
Figure 1: Poison control center (PCC)-emergency department (ED) information exchange events and relevant Consolidated Clinical Document Architecture (C-CDA) documents.
523
RESEARCH AND APPLICATIONS
Mappings from PCC–ED HIE information types to C-CDA sections All PCC–ED information types could be mapped to at least one C-CDA section (table 3). However, not all mappings were exact. Three types of mapping were identified: perfect match, subset match, and multiple match. In perfect matches the information type was identical to the C-CDA section, such as in diagnostic test results. Subset matches consisted of cases in which the
Del Fiol G, et al. J Am Med Inform Assoc 2015;22:519–528. doi:10.1136/amiajnl-2014-003127, Research and applications
Table 2: Mapping between PCC–ED HIE events and C-CDA documents
RESEARCH AND APPLICATIONS
PCC–ED HIE event
Sender
Receiver
C-CDA document
Intent
PCC notifies ED of incoming patient
PCC
ED
Consultation Note
A summary of the patient’s complaint/exposure and an initial set of recommendations help the ED personnel prepare for the patient’s arrival.
Request for consultation
ED
PCC
Progress Report
An initial patient status report establishes an ongoing PCC–ED consultation. The Progress Report C-CDA includes patient assessment information that supports the poison control specialist’s review of the case, provision of more detailed or updated recommendations, and entry of data into the NPDS.
Patient status update
ED
PCC
Progress Report
The ED provides periodic patient status updates, including vital signs, laboratory test results, and medications. This allows the poison control specialists to reassess the patient, update their recommendations if needed, and enter data into the NPDS.
Treatment recommendations
PCC
ED
Consultation Note
The PCC specialists provide updated recommendations for the patient’s treatment. This event is equivalent to a specialist consultation, in which the specialist provides a consultation note to the referring clinician.
Patient discharged
ED
PCC
Discharge Summary
The ED submits a summary of the patient’s visit and discharge disposition. This allows the PCC to close the case, enter data into the NPDS and, if applicable, prepare a case notification report for public health purposes.
C-CDA, Consolidated Clinical Document Architecture; ED, emergency department; HIE, health information exchange; NPDS, National Poison Data System; PCC, poison control center.
desired information type may be found in a C-CDA section that also includes other kinds of information unrelated to the information type. For example, the information type patient-reported symptoms was matched to the C-CDA history of present illness section, which contains a complete patient history that includes symptoms as well as other information on the patient’s history. Multiple matches occurred for information types that mapped to more than one C-CDA section. For example, the information type absence of clinical effects may be found in three sections of the C-CDA History & Physical Note: review of systems, history of present illness, and physical exam. Subset and multiple matches were typically related to different granularity between information types and C-CDA sections. One important issue was that although PCCs collect information about the poison exposure in a structured coded form, according to the NPDS requirements, this information cannot be represented in the same structured format in a C-CDA template. In the absence of a specific template, poisoning exposure information was mapped to the history of present illness section, which only includes information in narrative form.
with the potential to improve patient outcomes, ED care processes, and public health reporting. The results of our assessment will be used to guide the design and development of a scalable and standards-based electronic PCC–ED HIE process. Our work has two major strengths. First, rather than proposing a new standard specific for PCC–ED HIE, we leveraged existing standards that are being rapidly adopted, particularly as a result of the EHR Meaningful Use certification process in the USA.15 Second, the PCC–ED information types were identified through rigorous analysis of PCC–ED telephone conversations. However, our analysis should be considered with caution since we did not assess issues related to the implementation of the C-CDA standard for PCC–ED HIE. First, the proposed mappings have not been validated in actual system implementations. Therefore, our results should serve as a guideline and not as a final standard specification for PCC–ED HIE. Second, variations in data expression and errors in current implementations of the C-CDA standard have been identified as barriers to semantic interoperability. These variations may compromise the broad generalizability of a C-CDA-based approach to PCC–ED HIE.16 Last, our analysis did not cover the dynamic messaging component of PCC–ED HIE, which could be accomplished through the adoption of other Meaningful Use standards, such as the Nationwide Health Information Network (NwHIN) standards.17 Overall, PCC–ED HIE events mapped well to C-CDA documents and PCC–ED information types mapped reasonably well
DISCUSSION To our knowledge, this is the first work to systematically identify and assess clinical data standards that can support bidirectional HIE between PCCs and EDs. This kind of HIE is a promising approach to improve the safety and efficiency of the widely used telephone-based PCC–ED communication process,
524
Del Fiol G, et al. J Am Med Inform Assoc 2015;22:519–528. doi:10.1136/amiajnl-2014-003127, Research and applications
Table 3: Mapping between PCC–ED information types and C-CDA sections Poisoning information type
C-CDA section(s)
Mapping type
PCC to ED: PCC notifies ED of incoming patient and poisoning consultation (C-CDA: Consultation Note) Exposure information*
History of present illness
Subset
Medical history
History of past illness
Subset
Patient medications
Medications
Perfect
Chief complaint/reason for visit
Chief complaint
Perfect
Absence of clinical effects
History of present illness
Subset
Caller reported mental status
History of present illness
Subset
Caller reported symptoms
History of present illness
Subset
Unrelated symptoms
History of present illness
Subset
†
Subset
PCC recommendations for treatment and discharge parameters
Plan of care
Assessment
Subset
Toxicity levels
Assessment
Subset
Clinical effects of substance
Assessment
Subset
Worst case scenario
Assessment
Subset
ED to PCC: request for consultation (C-CDA: History and Physical) Exposure information
History of present illness
Subset
Medical history
History of past illness
Subset
Patient medications
Medications
Perfect
Chief complaint/reason for visit
Chief complaint
Perfect
Absence of clinical effects
Review of systems, history of present illness, and physical exam
Multiple
Mental status assessment
History of present illness
Subset
Patient reported symptoms
History of present illness
Subset
Physical exam findings
Physical exam
Perfect
Unrelated symptoms
History of present illness, review of systems
Multiple
Vital signs
Vital signs
Perfect
Confirmation that treatment was given
Procedures, medications
Multiple
Patient status
Assessment
Subset
Plan of care
Plan of care
Perfect
Diagnostic test results
Results
Perfect
Time next labs/tests ordered
Plan of care
Subset
Time test performed/lab drawn
Plan of care
Subset
ED to PCC: patient status update (C-CDA: Progress Report) Chief complaint/reason for visit
Reason for visit
Perfect
Absence of clinical effects
Objective, subjective, review of systems
Multiple (continued)
525
RESEARCH AND APPLICATIONS
Toxic dose
Del Fiol G, et al. J Am Med Inform Assoc 2015;22:519–528. doi:10.1136/amiajnl-2014-003127, Research and applications
Table 3: Continued Poisoning information type
RESEARCH AND APPLICATIONS
C-CDA section(s)
Mapping type
Mental status assessment
Objective
Subset
Patient reported symptoms
Subjective, review of systems
Multiple
Physical exam findings
Physical exam
Perfect
Unrelated symptoms
Subjective, review of systems
Multiple
Vital signs
Vital signs
Perfect
Confirmation treatment was given
Procedures, medications
Multiple
Patient status
Assessment
Subset
Plan of care
Plan of care
Perfect
Diagnostic test results
Results
Perfect
Time next labs/tests ordered
Plan of care
Subset
Time test performed/lab drawn
Plan of care
Subset
ED to PCC: patient discharged (C-CDA: Discharge Summary) Exposure information
History of present illness
Subset
Medical history
History of past illness
Subset
Patient medications
Hospital discharge medications
Perfect
Chief complaint/reason for visit
Reason for visit
Perfect
Absence of clinical effects
History of present illness, review of systems
Multiple
Mental status assessment
History of present illness
Subset
Patient reported symptoms
History of present illness
Subset
Unrelated symptoms
History of present illness
Subset
Vital signs
Vital signs
Perfect
Confirmation treatment was given
Hospital discharge medications, procedures
Multiple
Patient discharge medications
Hospital discharge medications
Perfect
Patient status
History of present illness
Subset
Plan of care
Plan of care
Perfect
Diagnostic test results
Hospital discharge studies summary
Perfect
Time next labs/tests ordered
Plan of care
Subset
Time test performed/lab drawn
Plan of care
Subset
*PCCs capture exposure information according to a structured format that includes the NPDS data elements. Since the history of present illness is a narrative template, the information in NPDS format would need to be converted to narrative form in the Consultation Note and History & Physical Note documents. † The Plan of Care template is structured according to items such as Orders and Procedures. C-CDA, Consolidated Clinical Document Architecture; ED, emergency department; NPDS, National Poison Data System; PCC, poison control center.
to C-CDA sections. The main limitation was the lack of a structured CDA template that matches the NPDS format for poisoning exposure information, resulting in two problems: (1) structured poison exposure information collected at PCCs would need to be transmitted in narrative form to EDs; and (2)
poison exposure information collected at the EDs in narrative form would still need to be manually entered by PCC specialists in structured form according to the NPDS. To address these issues, we are planning to work with HL7 to develop a structured template to describe poison exposure according to the
526
Del Fiol G, et al. J Am Med Inform Assoc 2015;22:519–528. doi:10.1136/amiajnl-2014-003127, Research and applications
FUNDING
NPDS format. In ongoing work, we are mapping clinical terminologies to the NPDS. Preliminary results of this work are reported elsewhere.11 Standardization of information sharing between PCCs and EDs, enabled by the HL7 C-CDA standard, holds important and positive implications for public health. Currently, PCCs collect data for the NPDS. This national surveillance system only includes cases reported to the PCC. However, many poison exposures are treated in EDs without PCC consultation. We are designing an HIE process that entails universal poison exposure case notification to PCCs based upon triggers (figure 1, arrow (B)). Essentially, the PCC will be notified of all potential poison exposure events presenting to participating EDs, greatly enhancing their ability to effectively surveil and respond to poisoning as a population health concern. Further, the quality of the data available to the PCC will be much higher, enabling more sophisticated approaches to surveillance and analysis. Standardization of information sharing may also ameliorate many key safety vulnerabilities in the process of caring for individual patients, including poor information quality, information loss, and lack of information availability to multiple care team members. High quality, structured information can be presented in a way that enhances decision-making while reducing interruption, because it can be parsed and managed in ED information and tracking systems. Detailed information can be made accessible to the entire care team, and care providers can be notified of new information without the intrusiveness of a telephone call. This enables an efficient flow of information between EDs and PCCs, minimizing interruption and maximizing the timely availability of relevant information at the point of clinical decision-making.
This work was supported by the US Department of Health and Human Services, Agency for Healthcare Research and Quality, grant number R01 HS21472-01A1.
COMPETING INTERESTS None.
PROVENANCE AND PEER REVIEW Not commissioned; externally peer reviewed.
REFERENCES
CONCLUSION We identified and assessed clinical data standards to support the electronic exchange of information between PCCs and EDs. This process could be supported by four clinical documents from the HL7 C-CDA standard: History & Physical Note, Consultation Note, Progress Note, and Discharge Summary. Except for information related to the poison exposure, information types that are commonly exchanged between PCCs and EDs can be reasonably well represented within these C-CDA documents. Therefore, standards-based HIE between PCCs and EDs appears to be feasible given a set of standards that are required for EHR certification in the USA. However, the current standards framework could be improved with a CDA template that captures information exposure according to the NPDS data elements. In addition, the proposed approach still needs to be validated in actual system implementations.
CONTRIBUTORS All authors provided substantial contributions to the conception of the work; development, analysis, and interpretation of information mappings; and drafting and revising of the manuscript. All authors approved the final version to be published and agree to be accountable for all aspects of the work.
527
RESEARCH AND APPLICATIONS
1. Fingerhut LA, Anderson RN. The three leading causes of injury mortality in the United States, 1999–2005. NCHS Health E-Stats: Centers for Disease Control & Prevention, 2008. 2. Administration SAaMHS. Drug abuse warning network, 2011: national estimates of drug-related emergency department visits. Rockville, MD: Substance Abuse and Mental Health Services Administration, 2013. 3. Web-based Injury Statistics Query and Reporting System (WISQARS) [program]. 2013. http://www.cdc.gov/injury/ wisqars/ 4. Bond GR, Woodward RW, Ho M. The growing impact of pediatric pharmaceutical poisoning. J Pediatr. 2012;160: 265–70.e1. 5. Cummins MR, Crouch B, Gesteland P, et al. Inefficiencies and vulnerabilities of telephone-based communication between U. S. poison control centers and emergency departments. Clin Toxicol. 2013;51:435–43. 6. Cummins MR, Crouch BI, Gesteland P, et al. Electronic information exchange between emergency departments and poison control centers: a Delphi study. Clin Toxicol. 2012;50:503–13. 7. Cummins MR. Electronic Exchange of Poisoning Information (1 R01 HS021472–01A1): U.S. Agency for Healthcare Research & Quality). 2013. 8. HL7 Implementation Guide for CDA Release 2: IHE Health Story Consolidation, Release 1.1—US Realm: Health Level Seven International. 2012. 9. Mowry JB, Spyker DA, Cantilena LRJr, et al. 2012 Annual Report of the American Association of Poison Control Centers’ National Poison Data System (NPDS): 30th Annual Report. Clin Toxicol. 2013;51:949–1229. 10. Peters MS, Poynton MR, Ellington L, et al. Use of standardized terminologies in the poison control setting: an analysis. 10th International Congress on Nursing Informatics. 2009; Helsinki, Finland. 11. Cummins MR, Doing-Harris KM, Passman J, et al. Automated mapping of NPDS data elements to the UMLS Metathesaurus. Proceedings of the AMIA Annual Fall Symposium. Washington: DC: AMIA Fall Symposium, 2013:267. 12. Cummins MR, Crouch BI, Del Fiol G, et al. Information requirements for health information exchange supported
Del Fiol G, et al. J Am Med Inform Assoc 2015;22:519–528. doi:10.1136/amiajnl-2014-003127, Research and applications
practices: United States, 2001–2013. NCHS Data Brief. 2014;(143):1–8. 16. D’Amore JD, Mandel JC, Kreda DA, et al. Are Meaningful Use Stage 2 certified EHRs ready for interoperability? Findings from the SMART C-CDA Collaborative. J Am Med Inform Assoc. 2014;21:1060–8. 17. HealthIT.gov. Nationwide Health Information Network (NwHIN). http://www.healthit.gov/policy-researchers-implementers/nationwide-health-information-network-nwhin 18. American Association of Poison Control Centers. National C Reference Manual Part 2— Poison Data System (NPDS)V System Information Manual, 2009:156.
communication between emergency departments and poison control centers. Proceedings of the AMIA Annual Fall Symposium Washington, DC, in press, 2014. 13. Dolin RH, Alschuler L, Boyer S, et al. HL7 Clinical Document Architecture, Release 2. J Am Med Inform Assoc. 2006;13: 30–9. 14. Electronic Health Record Incentive Program-Stage 2. Regulation Identification Number 0938-AQ84: United States Department of Health and Human Services. Centers for Medicare and Medicaid Services, 2012. 15. Hsiao CJ, Hing E. Use and characteristics of electronic health record systems among office-based physician
AUTHOR AFFILIATIONS .................................................................................................................................................... 1
3
Department of Biomedical Informatics, University of Utah, Salt Lake City, Utah, USA
College of Pharmacy, University of Utah, Salt Lake City, Utah, USA
2
RESEARCH AND APPLICATIONS
College of Nursing, University of Utah, Salt Lake City, Utah, USA
528
Data standards to support health information exchange between poison control centers and emergency departments
RECEIVED 9 July 2014 REVISED 3 October 2014 ACCEPTED 12 October 2014 PUBLISHED ONLINE FIRST 23 October 2014
Guilherme Del Fiol1, Barbara Insley Crouch2, Mollie R Cummins1,3
ABSTRACT ....................................................................................................................................................
.................................................................................................................................................... Key words: poison control, clinical decision-making, health information exchange, electronic health records
INTRODUCTION
routinely call poison centers for consultation. In a typical patient consultation, both PCC specialists and ED providers place several phone calls throughout the ED admission. In these phone calls, a significant amount of information is exchanged such as a case summary, description of the exposure, the patient’s current status, diagnostic test results, and recommendations for treatment. In previous work, we identified a number of inefficiencies and safety vulnerabilities in the telephone-based ED–PCC communication process.5 These inefficiencies and safety vulnerabilities included repetition of redundant clinical information, vague or ambiguous communication of clinical observations and diagnostic test results, and difficulty establishing telephone communication due to provider unavailability. Through a nationwide Delphi study, we also identified widespread support for a health information exchange (HIE)-supported communication process.6 Based upon the results of that study, we propose a new process in which HIE does not entirely eliminate
Poisoning deaths have increased dramatically over the last 10 years, surpassing firearm deaths in 2004 and motor vehicle traffic deaths in 2011. The dramatic increase in deaths is largely due to medication overdose, especially opioid pain medications.1 Emergency department (ED) visits for poisoning have also dramatically increased during the same period.2–4 US poison control centers (PCCs) play a critical role in the management of poison exposures. These centers are staffed 24 h a day by specialists in poison information who are health professionals specially trained in toxicology. These specialists assess poison exposures over the telephone and provide treatment recommendations. This service is available to both healthcare providers and the general public. PCCs also collect detailed information about exposure cases in order to provide public health services and poison surveillance. Given the sheer diversity of potential poisons and the complexity of managing unusual and multi-substance exposures, ED care providers
Correspondence to Dr Guilherme Del Fiol, Department of Biomedical Informatics, University of Utah, 421 Wakara Way, Suite 140, Salt Lake City, UT 84108, USA; E-mail: [email protected] C The Author 2014. Published by Oxford University Press on behalf of the American Medical Informatics Association. V All rights reserved. For Permissions, please email: [email protected] For numbered affiliations see end of article.
519
RESEARCH AND APPLICATIONS
Objective Poison control centers (PCCs) routinely collaborate with emergency departments (EDs) to provide care for poison-exposed patients. During this process, a significant amount of information is exchanged between EDs and PCCs via telephone, leading to important inefficiencies and safety vulnerabilities. In the present work, we identified and assessed a set of data standards to enable a standards-based health information exchange process between EDs and PCCs. Materials and methods Based on a reference model for PCC–ED health information exchange, we (1) mapped PCC–ED information exchange events to clinical documents specified in the Health Level Seven (HL7) Consolidated Clinical Document Architecture (C-CDA) Standard, and (2) mapped information types routinely exchanged in PCC–ED telephone conversations to C-CDA sections. Results Four C-CDA document types were necessary to support the PCC–ED information exchange process: History & Physical Note, Consultation Note, Progress Note, and Discharge Summary. Information types that are commonly exchanged between PCCs and EDs can be reasonably well represented within these C-CDA documents. Conclusions A standards-based health information exchange process between PCCs and EDs appears to be feasible given a set of clinical data standards that are required for EHR certification in the USA, although the proposed approach still needs to be validated in actual system implementations. Such a process has the potential to improve the safety and efficiency of PCC–ED communication, ultimately resulting in improved patient care outcomes.
Del Fiol G, et al. J Am Med Inform Assoc 2015;22:519–528. doi:10.1136/amiajnl-2014-003127, Research and applications
standardized terminologies.10,11 Consequently, NPDS data are valuable for poisoning surveillance but are poorly interoperable from a clinical informatics perspective.
RESEARCH AND APPLICATIONS
telephone consultations, but supports and reduces the need for synchronous telephone communication. HIE is potentially valuable in supporting continuity of care as patients transition from one healthcare setting to another, especially when transitions of care occur across healthcare organizations without interoperable information systems. Currently, we are developing a scalable, standards-based approach to HIE between EDs and PCCs as part of an AHRQ (US Agency for Healthcare Research and Quality)-funded study to evaluate the effect of HIE-supported communication on workflow, communication, efficiency, and utilization.7 This type of HIE is distinctive in that it will involve bidirectional information sharing and communication in support of an emergency healthcare event, as it unfolds. This will entail multiple HIE messages that support real-time collaboration and decisionmaking. The goal of the present work was to identify and assess a set of data standards to support dynamic electronic information sharing between EDs and PCCs. We focused on standards that are required for EHR Meaningful Use certification in the USA, specifically the Health Level Seven (HL7) Consolidated Clinical Document Architecture (C-CDA) Standard.8
PCC–ED information types In previous work, we analyzed call recordings corresponding to a random sample of 120 poison exposure cases entailing telephone communication between an ED and the Utah Poison Control Center.5,12 Using qualitative methodology, we identified 52 unique information types exchanged during ED–PCC telephone communication. A subset of these information types that we selected for mapping to the C-CDA standard are described in table 1. These information types, derived from real-world communication between EDs and PCCs and well-validated, encapsulate the focused subset of information currently used to support collaboration.12 Several of the information types closely correspond to NPDS data elements, since communication partly serves to support collection of information for entry into the poison center information system, which collects NPDS data elements. However, the information types used for mapping reflect the overall ED–PCC communication that currently takes place in support of both patient care and population health.
BACKGROUND AND SIGNIFICANCE Collaboration and communication between PCCs and EDs Collaboration between an ED and a PCC occurs in three phases, described in detail in Cummins et al,5 and can be initiated by either party. In some cases, the PCC refers a patient to an ED for further evaluation and treatment. In other cases, a patient presents to an ED without contacting the PCC, and collaboration is initiated by an ED physician, nurse, or other staff member. Over the telephone, the PCC specialist assesses the poison exposure, provides consultation, and recommends a plan of care. Communication is ongoing as the case unfolds. The PCC calls periodically to request information on the patient’s clinical condition, diagnostic test results, treatment, and plan of care, updating treatment recommendations as necessary. The purpose of ongoing telephone communication is twofold: (1) to support ongoing emergency treatment for poison exposure; and (2) to support the collection of case data for both local (state and local health departments) and national poisoning surveillance (National Poison Data System, NPDS).9 However, reliance upon telephone communication leads to inefficiencies and safety vulnerabilities.5
Consolidated Clinical Document Architecture Standard The HL7 CDA standard provides a framework for representing various kinds of clinical documents, such as progress notes, discharge summaries, and imaging reports.13 CDA documents consist of a header and a body. The header includes document metadata, such as document type, patient demographics, care provider and/or healthcare institution responsible for generating the document, and document timestamp. The body contains the patient’s clinical data, which are organized into multiple sections. Each section is modeled based on standard reusable templates, such as plan of care, problems, medications, and results. Therefore, the CDA body is constructed with modular blocks of data, with standard templates providing the building blocks. The C-CDA standard is an HL7 implementation guide that builds upon CDA release 2.0. It defines nine standard document types for different clinical use cases: Continuity of Care Document, Consultation Note, Discharge Summary, Procedure Note, Diagnostic Imaging Reports, History and Physical, Operative Note, Progress Note, and Unstructured Documents.8 The C-CDA standard is required for EHR certification in the USA EHR Meaningful Use program.14 Since the C-CDA standard was designed to cover a broad range of clinical use cases, it is a reasonable choice to support PCC–ED HIE.
Data elements for poison exposure information Using a proprietary, purpose-built information system, the PCC collects structured and narrative data describing each poison exposure case. The structured data largely consist of NPDS data elements. The NPDS data elements allow for the collection of highly detailed information, in both structured and narrative form, about the nature and circumstances of each poison exposure. The NPDS data elements are not represented in any standard information model and coded data are not represented in standard controlled terminologies. In previous work, we identified that these data elements only partially map to
MATERIALS AND METHODS Our approach entailed three steps. First, we designed a reference model for PCC–ED HIE events. This model was informed by our previous work, in which we systematically observed and analyzed PCC–ED phone conversations.5 Second, we mapped each PCC–ED HIE event to a set of candidate C-CDA document types. This mapping was informed by the document definitions
520
Duration of exposure (acute vs chronic) Narrative information about the poison exposure scenario specifically relevant to determining the certainty of ingestion/exposure, and general information helpful in constructing a clinical picture of the patient and treatment plan General grouping or characterization of a poison (eg, b-blocker) General characteristics of a poison (eg, half-life, mechanism of action, peak effect) Generic name of a poison that is also a therapeutic agent Dosage or amount of substance (eg, 500 mg metformin). Brand name of a poison that is also a therapeutic agent Informal description ranging from general drug class (eg, antiarrhythmic) to intended purpose (eg, a chemical used to clean carburetors) For example, tablet, pill, powder, lozenges For example, extended release vs rapid release Narrative describing reasoning of a PCC specialist in identifying a substance based on characteristics of the poison (eg, ‘A blue pill in that shape can be a few different things, but it is probably Viagra’) Name of a non-pharmacological substance For example, solid, inhalant, liquid Amount of time elapsed since the initial poison exposure (eg, ‘It has been about 5 h since she took the pills’)
Chronicity
Establishing background/certainty
Substance class
Substance information
Substance name (generic)
Substance amount
Substance name (brand)
Substance description
Substance form
Substance formulation
Substance identification rationale
Substance–non-pharmacological
Substance type
Time since ingestion
Information about the patient’s past medical history Medications that the patient currently takes at home
Medical history
Patient medications
(continued)
The level of certainty about the precise formulation, based on the subjective/objective information provided in a circumstance where the poison is a therapeutic agent
Certainty of formulation
RESEARCH AND APPLICATIONS
521
Patient health history
For example, polypharmacy, accidental exposure, overdose
Exposure type
Exposure information
Table 1: Information types identified in an observational study of ED–PCC telephone communication* Del Fiol G, et al. J Am Med Inform Assoc 2015;22:519–528. doi:10.1136/amiajnl-2014-003127, Research and applications
Signs observed by ED healthcare providers Healthcare provider or PCC assessment that a symptom is unrelated or likely unrelated to the poison exposure. For example, a patient may exhibit tremors related to underlying Parkinson’s disease, which are unrelated to an acute narcotics overdose Information describing the patient’s blood pressure, heart rate, and/or respiratory rate along with the date and time the vital signs were obtained
Physical exam findings
Unrelated symptoms
Vital signs
522
Specific dose or amount of exposure at which a substance becomes toxic Circulating blood level of a substance considered toxic Potential or expected clinical effects of a given poison exposure Description of the most harmful clinical effects and poorest outcome that a patient might experience (a description of risk)
Toxic dose
Toxicity levels
Clinical effects of substance
Worst case scenario
Medications prescribed for patient at time of discharge from the ED ED description of the patient’s clinical condition, particularly whether clinical effects of exposure are observed ED healthcare provider plan of care for the patient (eg, treatments, procedures, length of stay, parameters for discharge) Inquiry about results of diagnostic tests (eg, laboratory, ECG, imaging) Information about the ED’s planned or recommended timing of subsequent diagnostic testing (eg, ‘We’ll get another level at 4 h [after ingestion]’) Information about the timing of a treatment or a laboratory test, usually in relation to the time of ingestion (eg, ‘That level was drawn 2 h after ingestion’) Information that a treatment, directly related to the poison exposure, was administered
Patient discharge medications
Patient status
Plan of care
Diagnostic test results
Time next laboratory tests will be ordered
Time laboratory test was performed/drawn
Treatment performed
*Adapted from Cummins et al. One goal of telephone communication between EDs and PCCs is the collection of information for entry into the NPDS. Consequently, some information types resemble NPDS data elements. However, NPDS data elements are defined differently, according to a published coding manual.18 ED, emergency department; NPDS, National Poison Data System; PCC, poison control center.
12
ED staff member indicates whether or not a PCC recommended treatment has been administered to a patient
Confirmation that treatment was given
ED treatment/management information
PCC recommendations for treatment and/or duration of direct observation prior to discharge from the ED
PCC recommendations for treatment and discharge parameters
PCC recommendations and toxicology information
Information about the patient’s mental status Symptoms as reported by the patient or caller to the PCC
Caller reported symptoms
The absence of any signs or symptoms attributable to the poison exposure
Absence of clinical effects
Mental status
Patient’s chief complaint upon presenting to the emergency department or calling the PCC
Chief complaint/reason for visit
Subjective and objective information
RESEARCH AND APPLICATIONS
Table 1: Continued
Del Fiol G, et al. J Am Med Inform Assoc 2015;22:519–528. doi:10.1136/amiajnl-2014-003127, Research and applications
Del Fiol G, et al. J Am Med Inform Assoc 2015;22:519–528. doi:10.1136/amiajnl-2014-003127, Research and applications
provided in the C-CDA specification. Next, we mapped each PCC–ED information type to C-CDA sections contained in the documents identified in the previous step. This mapping was informed by the section and template definitions and examples provided in the C-CDA specification. We excluded information types that constituted questions or verbal requests for information, specific to telephone-based communication (eg, confirmation that the patient has arrived at the ED, inquiry regarding the patient’s discharge or transfer status), and information types routinely included in C-CDA headers. Last, for each PCC–ED HIE event, we selected the C-CDA document whose sections best matched the PCC–ED information types. Two investigators with expertise in HL7 standards (GDF) and PCC–ED information exchange (MRC) independently completed mappings. We reconciled the independent mappings iteratively and through consensus with a poison control expert (BIC).
different ways: (A) The PCC receives a phone call from a patient and refers that patient to the ED by sending an initial C-CDA Consultation Note; or (B) the ED receives a poison-exposed patient and requests consultation from the PCC through a C-CDA History & Physical Note. Once a case consultation is established, a set of iterative events take place depending on the patient’s progress: the ED submits periodic patient status updates to the PCC using the Progress Note C-CDA (C); and the PCC submits any updated management recommendations to the ED using the Consultation Note C-CDA (D). Finally, when the patient is transferred or discharged, the ED sends a C-CDA Discharge Summary document to the PCC, closing the case (E). Table 2 describes each of the PCC–ED HIE events.
RESULTS Reference model for PCC–ED HIE events with mappings to C-CDA document types Figure 1 depicts the HIE events necessary to support ED–PCC communication and collaboration. The process may begin in two
Figure 1: Poison control center (PCC)-emergency department (ED) information exchange events and relevant Consolidated Clinical Document Architecture (C-CDA) documents.
523
RESEARCH AND APPLICATIONS
Mappings from PCC–ED HIE information types to C-CDA sections All PCC–ED information types could be mapped to at least one C-CDA section (table 3). However, not all mappings were exact. Three types of mapping were identified: perfect match, subset match, and multiple match. In perfect matches the information type was identical to the C-CDA section, such as in diagnostic test results. Subset matches consisted of cases in which the
Del Fiol G, et al. J Am Med Inform Assoc 2015;22:519–528. doi:10.1136/amiajnl-2014-003127, Research and applications
Table 2: Mapping between PCC–ED HIE events and C-CDA documents
RESEARCH AND APPLICATIONS
PCC–ED HIE event
Sender
Receiver
C-CDA document
Intent
PCC notifies ED of incoming patient
PCC
ED
Consultation Note
A summary of the patient’s complaint/exposure and an initial set of recommendations help the ED personnel prepare for the patient’s arrival.
Request for consultation
ED
PCC
Progress Report
An initial patient status report establishes an ongoing PCC–ED consultation. The Progress Report C-CDA includes patient assessment information that supports the poison control specialist’s review of the case, provision of more detailed or updated recommendations, and entry of data into the NPDS.
Patient status update
ED
PCC
Progress Report
The ED provides periodic patient status updates, including vital signs, laboratory test results, and medications. This allows the poison control specialists to reassess the patient, update their recommendations if needed, and enter data into the NPDS.
Treatment recommendations
PCC
ED
Consultation Note
The PCC specialists provide updated recommendations for the patient’s treatment. This event is equivalent to a specialist consultation, in which the specialist provides a consultation note to the referring clinician.
Patient discharged
ED
PCC
Discharge Summary
The ED submits a summary of the patient’s visit and discharge disposition. This allows the PCC to close the case, enter data into the NPDS and, if applicable, prepare a case notification report for public health purposes.
C-CDA, Consolidated Clinical Document Architecture; ED, emergency department; HIE, health information exchange; NPDS, National Poison Data System; PCC, poison control center.
desired information type may be found in a C-CDA section that also includes other kinds of information unrelated to the information type. For example, the information type patient-reported symptoms was matched to the C-CDA history of present illness section, which contains a complete patient history that includes symptoms as well as other information on the patient’s history. Multiple matches occurred for information types that mapped to more than one C-CDA section. For example, the information type absence of clinical effects may be found in three sections of the C-CDA History & Physical Note: review of systems, history of present illness, and physical exam. Subset and multiple matches were typically related to different granularity between information types and C-CDA sections. One important issue was that although PCCs collect information about the poison exposure in a structured coded form, according to the NPDS requirements, this information cannot be represented in the same structured format in a C-CDA template. In the absence of a specific template, poisoning exposure information was mapped to the history of present illness section, which only includes information in narrative form.
with the potential to improve patient outcomes, ED care processes, and public health reporting. The results of our assessment will be used to guide the design and development of a scalable and standards-based electronic PCC–ED HIE process. Our work has two major strengths. First, rather than proposing a new standard specific for PCC–ED HIE, we leveraged existing standards that are being rapidly adopted, particularly as a result of the EHR Meaningful Use certification process in the USA.15 Second, the PCC–ED information types were identified through rigorous analysis of PCC–ED telephone conversations. However, our analysis should be considered with caution since we did not assess issues related to the implementation of the C-CDA standard for PCC–ED HIE. First, the proposed mappings have not been validated in actual system implementations. Therefore, our results should serve as a guideline and not as a final standard specification for PCC–ED HIE. Second, variations in data expression and errors in current implementations of the C-CDA standard have been identified as barriers to semantic interoperability. These variations may compromise the broad generalizability of a C-CDA-based approach to PCC–ED HIE.16 Last, our analysis did not cover the dynamic messaging component of PCC–ED HIE, which could be accomplished through the adoption of other Meaningful Use standards, such as the Nationwide Health Information Network (NwHIN) standards.17 Overall, PCC–ED HIE events mapped well to C-CDA documents and PCC–ED information types mapped reasonably well
DISCUSSION To our knowledge, this is the first work to systematically identify and assess clinical data standards that can support bidirectional HIE between PCCs and EDs. This kind of HIE is a promising approach to improve the safety and efficiency of the widely used telephone-based PCC–ED communication process,
524
Del Fiol G, et al. J Am Med Inform Assoc 2015;22:519–528. doi:10.1136/amiajnl-2014-003127, Research and applications
Table 3: Mapping between PCC–ED information types and C-CDA sections Poisoning information type
C-CDA section(s)
Mapping type
PCC to ED: PCC notifies ED of incoming patient and poisoning consultation (C-CDA: Consultation Note) Exposure information*
History of present illness
Subset
Medical history
History of past illness
Subset
Patient medications
Medications
Perfect
Chief complaint/reason for visit
Chief complaint
Perfect
Absence of clinical effects
History of present illness
Subset
Caller reported mental status
History of present illness
Subset
Caller reported symptoms
History of present illness
Subset
Unrelated symptoms
History of present illness
Subset
†
Subset
PCC recommendations for treatment and discharge parameters
Plan of care
Assessment
Subset
Toxicity levels
Assessment
Subset
Clinical effects of substance
Assessment
Subset
Worst case scenario
Assessment
Subset
ED to PCC: request for consultation (C-CDA: History and Physical) Exposure information
History of present illness
Subset
Medical history
History of past illness
Subset
Patient medications
Medications
Perfect
Chief complaint/reason for visit
Chief complaint
Perfect
Absence of clinical effects
Review of systems, history of present illness, and physical exam
Multiple
Mental status assessment
History of present illness
Subset
Patient reported symptoms
History of present illness
Subset
Physical exam findings
Physical exam
Perfect
Unrelated symptoms
History of present illness, review of systems
Multiple
Vital signs
Vital signs
Perfect
Confirmation that treatment was given
Procedures, medications
Multiple
Patient status
Assessment
Subset
Plan of care
Plan of care
Perfect
Diagnostic test results
Results
Perfect
Time next labs/tests ordered
Plan of care
Subset
Time test performed/lab drawn
Plan of care
Subset
ED to PCC: patient status update (C-CDA: Progress Report) Chief complaint/reason for visit
Reason for visit
Perfect
Absence of clinical effects
Objective, subjective, review of systems
Multiple (continued)
525
RESEARCH AND APPLICATIONS
Toxic dose
Del Fiol G, et al. J Am Med Inform Assoc 2015;22:519–528. doi:10.1136/amiajnl-2014-003127, Research and applications
Table 3: Continued Poisoning information type
RESEARCH AND APPLICATIONS
C-CDA section(s)
Mapping type
Mental status assessment
Objective
Subset
Patient reported symptoms
Subjective, review of systems
Multiple
Physical exam findings
Physical exam
Perfect
Unrelated symptoms
Subjective, review of systems
Multiple
Vital signs
Vital signs
Perfect
Confirmation treatment was given
Procedures, medications
Multiple
Patient status
Assessment
Subset
Plan of care
Plan of care
Perfect
Diagnostic test results
Results
Perfect
Time next labs/tests ordered
Plan of care
Subset
Time test performed/lab drawn
Plan of care
Subset
ED to PCC: patient discharged (C-CDA: Discharge Summary) Exposure information
History of present illness
Subset
Medical history
History of past illness
Subset
Patient medications
Hospital discharge medications
Perfect
Chief complaint/reason for visit
Reason for visit
Perfect
Absence of clinical effects
History of present illness, review of systems
Multiple
Mental status assessment
History of present illness
Subset
Patient reported symptoms
History of present illness
Subset
Unrelated symptoms
History of present illness
Subset
Vital signs
Vital signs
Perfect
Confirmation treatment was given
Hospital discharge medications, procedures
Multiple
Patient discharge medications
Hospital discharge medications
Perfect
Patient status
History of present illness
Subset
Plan of care
Plan of care
Perfect
Diagnostic test results
Hospital discharge studies summary
Perfect
Time next labs/tests ordered
Plan of care
Subset
Time test performed/lab drawn
Plan of care
Subset
*PCCs capture exposure information according to a structured format that includes the NPDS data elements. Since the history of present illness is a narrative template, the information in NPDS format would need to be converted to narrative form in the Consultation Note and History & Physical Note documents. † The Plan of Care template is structured according to items such as Orders and Procedures. C-CDA, Consolidated Clinical Document Architecture; ED, emergency department; NPDS, National Poison Data System; PCC, poison control center.
to C-CDA sections. The main limitation was the lack of a structured CDA template that matches the NPDS format for poisoning exposure information, resulting in two problems: (1) structured poison exposure information collected at PCCs would need to be transmitted in narrative form to EDs; and (2)
poison exposure information collected at the EDs in narrative form would still need to be manually entered by PCC specialists in structured form according to the NPDS. To address these issues, we are planning to work with HL7 to develop a structured template to describe poison exposure according to the
526
Del Fiol G, et al. J Am Med Inform Assoc 2015;22:519–528. doi:10.1136/amiajnl-2014-003127, Research and applications
FUNDING
NPDS format. In ongoing work, we are mapping clinical terminologies to the NPDS. Preliminary results of this work are reported elsewhere.11 Standardization of information sharing between PCCs and EDs, enabled by the HL7 C-CDA standard, holds important and positive implications for public health. Currently, PCCs collect data for the NPDS. This national surveillance system only includes cases reported to the PCC. However, many poison exposures are treated in EDs without PCC consultation. We are designing an HIE process that entails universal poison exposure case notification to PCCs based upon triggers (figure 1, arrow (B)). Essentially, the PCC will be notified of all potential poison exposure events presenting to participating EDs, greatly enhancing their ability to effectively surveil and respond to poisoning as a population health concern. Further, the quality of the data available to the PCC will be much higher, enabling more sophisticated approaches to surveillance and analysis. Standardization of information sharing may also ameliorate many key safety vulnerabilities in the process of caring for individual patients, including poor information quality, information loss, and lack of information availability to multiple care team members. High quality, structured information can be presented in a way that enhances decision-making while reducing interruption, because it can be parsed and managed in ED information and tracking systems. Detailed information can be made accessible to the entire care team, and care providers can be notified of new information without the intrusiveness of a telephone call. This enables an efficient flow of information between EDs and PCCs, minimizing interruption and maximizing the timely availability of relevant information at the point of clinical decision-making.
This work was supported by the US Department of Health and Human Services, Agency for Healthcare Research and Quality, grant number R01 HS21472-01A1.
COMPETING INTERESTS None.
PROVENANCE AND PEER REVIEW Not commissioned; externally peer reviewed.
REFERENCES
CONCLUSION We identified and assessed clinical data standards to support the electronic exchange of information between PCCs and EDs. This process could be supported by four clinical documents from the HL7 C-CDA standard: History & Physical Note, Consultation Note, Progress Note, and Discharge Summary. Except for information related to the poison exposure, information types that are commonly exchanged between PCCs and EDs can be reasonably well represented within these C-CDA documents. Therefore, standards-based HIE between PCCs and EDs appears to be feasible given a set of standards that are required for EHR certification in the USA. However, the current standards framework could be improved with a CDA template that captures information exposure according to the NPDS data elements. In addition, the proposed approach still needs to be validated in actual system implementations.
CONTRIBUTORS All authors provided substantial contributions to the conception of the work; development, analysis, and interpretation of information mappings; and drafting and revising of the manuscript. All authors approved the final version to be published and agree to be accountable for all aspects of the work.
527
RESEARCH AND APPLICATIONS
1. Fingerhut LA, Anderson RN. The three leading causes of injury mortality in the United States, 1999–2005. NCHS Health E-Stats: Centers for Disease Control & Prevention, 2008. 2. Administration SAaMHS. Drug abuse warning network, 2011: national estimates of drug-related emergency department visits. Rockville, MD: Substance Abuse and Mental Health Services Administration, 2013. 3. Web-based Injury Statistics Query and Reporting System (WISQARS) [program]. 2013. http://www.cdc.gov/injury/ wisqars/ 4. Bond GR, Woodward RW, Ho M. The growing impact of pediatric pharmaceutical poisoning. J Pediatr. 2012;160: 265–70.e1. 5. Cummins MR, Crouch B, Gesteland P, et al. Inefficiencies and vulnerabilities of telephone-based communication between U. S. poison control centers and emergency departments. Clin Toxicol. 2013;51:435–43. 6. Cummins MR, Crouch BI, Gesteland P, et al. Electronic information exchange between emergency departments and poison control centers: a Delphi study. Clin Toxicol. 2012;50:503–13. 7. Cummins MR. Electronic Exchange of Poisoning Information (1 R01 HS021472–01A1): U.S. Agency for Healthcare Research & Quality). 2013. 8. HL7 Implementation Guide for CDA Release 2: IHE Health Story Consolidation, Release 1.1—US Realm: Health Level Seven International. 2012. 9. Mowry JB, Spyker DA, Cantilena LRJr, et al. 2012 Annual Report of the American Association of Poison Control Centers’ National Poison Data System (NPDS): 30th Annual Report. Clin Toxicol. 2013;51:949–1229. 10. Peters MS, Poynton MR, Ellington L, et al. Use of standardized terminologies in the poison control setting: an analysis. 10th International Congress on Nursing Informatics. 2009; Helsinki, Finland. 11. Cummins MR, Doing-Harris KM, Passman J, et al. Automated mapping of NPDS data elements to the UMLS Metathesaurus. Proceedings of the AMIA Annual Fall Symposium. Washington: DC: AMIA Fall Symposium, 2013:267. 12. Cummins MR, Crouch BI, Del Fiol G, et al. Information requirements for health information exchange supported
Del Fiol G, et al. J Am Med Inform Assoc 2015;22:519–528. doi:10.1136/amiajnl-2014-003127, Research and applications
practices: United States, 2001–2013. NCHS Data Brief. 2014;(143):1–8. 16. D’Amore JD, Mandel JC, Kreda DA, et al. Are Meaningful Use Stage 2 certified EHRs ready for interoperability? Findings from the SMART C-CDA Collaborative. J Am Med Inform Assoc. 2014;21:1060–8. 17. HealthIT.gov. Nationwide Health Information Network (NwHIN). http://www.healthit.gov/policy-researchers-implementers/nationwide-health-information-network-nwhin 18. American Association of Poison Control Centers. National C Reference Manual Part 2— Poison Data System (NPDS)V System Information Manual, 2009:156.
communication between emergency departments and poison control centers. Proceedings of the AMIA Annual Fall Symposium Washington, DC, in press, 2014. 13. Dolin RH, Alschuler L, Boyer S, et al. HL7 Clinical Document Architecture, Release 2. J Am Med Inform Assoc. 2006;13: 30–9. 14. Electronic Health Record Incentive Program-Stage 2. Regulation Identification Number 0938-AQ84: United States Department of Health and Human Services. Centers for Medicare and Medicaid Services, 2012. 15. Hsiao CJ, Hing E. Use and characteristics of electronic health record systems among office-based physician
AUTHOR AFFILIATIONS .................................................................................................................................................... 1
3
Department of Biomedical Informatics, University of Utah, Salt Lake City, Utah, USA
College of Pharmacy, University of Utah, Salt Lake City, Utah, USA
2
RESEARCH AND APPLICATIONS
College of Nursing, University of Utah, Salt Lake City, Utah, USA
528
