July 18, 2011
Medical Device Integration — More Than Meets the Eye
By Leslie Feldman
For The Record
Vol. 23 No. 13 P. 20
Because it generates overwhelming amounts of data, MDI is an issue HIM professionals need to address with hospital executives.
HIM experts forewarned the healthcare industry of an approaching HIPAA 5010/ICD-10/meaningful use perfect storm. What has gone largely unnoticed in the accompanying wave of EHR implementations is the oncoming tsunami of life-critical patient data generated by medical devices.
The U.S. medical device market, estimated to be $105.8 billion in 2011 by market research firm Espicom, is booming despite the overall economic downturn. The exponential growth is attributed in part to dramatic increases in the number of device types (eg, IV smart pumps, smart beds, bedside lab and vital sign monitors), patients being monitored across all venues of care, and devices deployed per patient in all care settings.
Most acutely ill patients can have a dozen medical monitors and devices continuously collecting massive volumes of life-critical data, according to Ann Farrell, BSN, principal of Farrell Associates LLC, a strategic HIT consulting firm currently focused on point-of-care solutions. Farrell adds that it’s not unusual for critical care RNs to document more than 1,000 data points per shift, including recording vital sign readings at increasingly frequent intervals.
Stillwater Medical Center, a 119-bed not-for-profit hospital in Oklahoma, recently updated its ICU patient monitoring equipment. “ICU nurses were spending a great deal of time manually documenting patient vital signs every 15 minutes and later reentering the information into the hospital information system,” says Chief Information Officer (CIO) Chris Roark. “The redundant documentation was highly inefficient and increased the likelihood of documentation errors. We wanted the nurses to focus on patients, not paperwork, so we decided to look for a solution to automate how vital sign data were captured and recorded in the ICU.”
Stillwater partnered with Accent on Integration, a systems integration firm focused on data sharing within the hospital and across the community, including the integration of medical device data with EHRs. Company Cofounder, President, and CEO Marc Andiel says having real-time, readily available data can make hospitals safer.
“Having data in silos or relying on data that are manually entered makes hospitals more prone to errors,” he notes.
Medical device integration (MDI) with clinical information systems (CIS), most notably EHRs, eliminates labor-intensive and error-prone transcription and the manual entry of monitor and device data into an EHR, Farrell says. With physiologic monitor data integration, RNs using critical thinking select data from continuous monitor readings that best represent a patient’s current status.
“MDI can drive tremendous value and thus far has really made a difference,” Roark says. “By harnessing the power of integration, Stillwater eradicated inefficient processes while improving patient care, workflow, and patient safety.”
Farrell says RNs filter medical device data prior to documenting in both paper-based and electronic processes. With MDI, device data are “imported” and validated by RNs prior to updating the EHR database.
“Some EHRs allow unvalidated data to be reviewed and accepted in the EHR flowsheet, while others require validation on a secondary screen prior to updating the EHR,” says Farrell, adding that without an RN medical device data quality assurance step, artifact and data outliers prevalent in critical care settings would be captured. This could result in patient harm if physicians in decision support use inaccurate data or EHR and surveillance systems trigger inappropriate and potentially contraindicated alerts, warnings, or conditional orders.
When indicated, Farrell says RNs associate discrete characteristics with individual data values, such as “pulse 80 weak” and “respirations 12 stertuous,” and document associated observations, such as “patient diaphoretic,” and responses to RN inquiries and interventions in the EHR. “Vital sign data can be a late-stage indicator of patient deterioration,” she explains. “Thus, RN assessment, including face-to-face observation (the primary reason for hospitalization), is critical for early detection of a change in patient status and timely intervention.”
Most lower acuity, subacute, and ambulatory patients are not monitored continuously, Farrell says. For these patients, RNs, or certified nursing assistants (CNAs) under the supervision of RNs, go bed to bed collecting vital sign and bedside lab data serially on a scheduled and “as-indicated” basis based on patient condition. Among the information gathered are data generated from diverse bedside devices such as glucometers, pulse oximeters, and medical monitors.
Each patient is associated with a device or devices. Farrell explains that values are typically presented to CNAs in batches for acceptance or reselection and acceptance prior to updating the EHR database. She says because patient assessment is beyond the scope of practice for CNAs, they do not append characteristics to values or chart other assessment information.
Peter Witonsky, president and chief sales officer of iSirona, an enterprise, vendor-agnostic, software-centric MDI company, stresses the importance of data validation. “From our perspective, clinician validation of data, or authentication prior to entry into the EMR, is an absolute must from a patient safety perspective,” he says. “The confirmation step provides clinicians the opportunity to quickly review device and ADT [admission, discharge, transfer] data, ensuring that the data are indeed associated to the correct patient and prevent errant or outlier device data from getting into the record.”
Cathleen Asch, president and CEO of Nuvon, an Internet-based, vendor-agnostic, enterprisewide and communitywide MDI firm, agrees: “There is no doubt that the accuracy of the information entered into EHRs is paramount to the quality of care predicated on the use of that EHR data. This has implications near-term for immediate decisions such as CPOE [computerized physician order entry] and long-term for outcomes analysis. Supporting clinicians with streamlined data verification options is critical to achieving this goal.”
Timely charting of patient status, including vital signs, is another element of data quality and best practice in both paper and electronic systems. Farrell says near-real-time charting of key patient assessment data becomes more critical as more physicians, clinicians, and care managers access an EHR remotely to review a patient’s condition and prescribe services. Physicians rely on up-to-date patient data to make safe, effective treatment decisions. EHR and surveillance systems also depend on these data for warnings, alerts, and conditional orders to be triggered at the appropriate time, not after the fact.
“The movement to real-time, point-of-care charting has been a culture shock for many nurses and a complex integration effort for CIOs and vendors,” Farrell notes.
“There is mounting pressure to chart ever-greater volumes of data at the point of care in near real time, which can be a daunting task,” she continues. “MDI mitigates the data collection process. Many RNs, stressed to the breaking point with care delivery burdens and increasingly onerous documentation requirements, clamor to recapture a half hour to hour of unproductive time to apply in direct patient care. We hear of a great number of critical care RNs, uncharacteristically in some cases, drawing a line in the sand in refusing to move from paper flowsheets and enter vital sign data into the EHR without MDI support.”
MDI approaches that automate flawed paper processes and workarounds will not improve—and can actually degrade—data quality and may introduce new problems, according to Farrell, adding that data integrity can be compromised in MDI in several ways, including the following:
• data on the wrong patient via inaccurate patient-to-device association (a reliance on a location-based solution) (Note: The Joint Commission has long established positive patient ID as a top safety goal. Bed- or location-based approaches are less safe than patient-centric solutions.)
• poorly authenticated physiologic monitoring device data.
• RN “back charting” for long periods as standard practice. Physicians performing CPOE may base decisions on old information when more recent data have been collected but not yet charted. While data charted long after the fact may later appear in the EHR at the time they were collected, the audit trail shows the actual time data were entered, presenting clinical and legal risks. Electronic back charting can mimic late-charting practices with paper processes.
It’s prudent for healthcare organizations to combine data quality and interoperability discussions, Farrell says. The IT industry has been focused on the integral role played by health data exchange in enabling meaningful use and meeting proposed health system reform measures. Farrell says MDI discussions often miss or undervalue concerns related to data quality, including the following:
• the accessibility, accuracy, and timeliness of data in existing paper, electronic, or hybrid systems;
• the cost and potentially fatal impact of using poor data in clinical decision support;
• the role of RN critical thinking in device-generated data quality assurance; and
• using vital signs as a “data set” as part of a larger RN patient assessment process.
“Several technical standards have been established for the transmission of data from medical devices,” says Brian McAlpine, director of strategic products at Capsule, an international vendor-agnostic, legacy enterprise MDI company. These include standards for the formatting of medical device data that are output to systems that collect and manage data.
Still, several important features, such as clinician workflow, are not addressed, says McAlpine. “Since many patients with medical devices are cared for in increasingly complex, chaotic environments, many more considerations come into play that all must come together, including patient safety and system reliability and supportability,” he says.
According to McAlpine, examples of important features not covered by technical standards include the following:
• High availability: data caching if a connection is broken or a network or a server goes down;
• Patient-to-device association: RNs must ensure each device is associated to the right patient;
• Real-time information at the bedside regarding connectivity and patient-to-device association;
• Filtering of data: Devices normally produce data at a much higher rate than an EMR can accept;
• Unit conversion: Devices from different vendors often use different units of measure. For example, temperatures produced in degrees Celsius may require conversion to degrees Fahrenheit.
“Standards will reduce the cost and complexity of device integration, but they are in no way the Holy Grail,” McAlpine says.
Asch says implementing MDI technology is a tall order. “The challenges of MDI adoption are multidimensional, with competing factors that must be solved satisfactorily for all,” she says. “Clinicians require seamless workflow integration that enhances rather than complicates their work; decision-support applications require real-time and synchronized data to ensure accurate and actionable information; IT needs flexible integration into a complex infrastructure that simplifies rather than compounds their work; and finance needs a cost-effective solution initially and over time that addresses the total cost of ownership.”
“Competing IT priorities and the associated budget concerns are challenges to widespread adoption of device connectivity,” says Witonsky. “That said, it quickly becomes a priority when hospitals witness how easily this technology impacts patient safety. Not only are nurses freed up to spend more time at the bedside, but patient outcomes improve when clinicians have faster access to accurate data in the EMR.”
As far as MDI’s role in the meaningful use equation, Farrell says medical device interoperability is currently included in the draft of stage 3 criteria. While MDI is not included in the stage 1 requirements or anticipated to be accelerated to stage 2, the stage 1 core set requirements for CPOE for medications and vital signs (manual data entry) are spotlighting existing vital sign data collection processes and the importance of data quality.
“Although not an early requirement for meaningful use, MDI can help healthcare organizations reach other components of the rule more safely and easier. Physicians cannot order certain tests, drugs, or treatments without access to a patient’s latest vital signs data,” McAlpine says.
The relationship between meaningful use and MDI is not being given its proper due, adds Andiel, noting that the “stage 1 vital signs requirement is only the tip of the iceberg.”
Recently, several HIT vendors performed research to evaluate the quality of vital signs data. In 2009, Intel Corporation and Motion Computing led a formal study on medical surgical units at an academic medical center transitioning to an EHR. Researchers reported “spot check” vital sign error rates of approximately 23% in the paper-based system and with the EHR using manual data entry.
Nurses working in the paper system jotted down device data at the bedside and later—sometimes much later—entered it in patients’ charts, creating two opportunities for potential errors to occur. After transitioning to EHRs, nurses memorized data collected at the bedside before entering them into the EHR via a workstation on wheels in the hallway, a common industry workaround. In a separate critical care case study sponsored by iSirona, Wise Regional Health System in Texas reported incidences of 12-hour lag times between when data were generated by devices and validated in the EHR prior to MDI.
The industry is experiencing an explosion of device use and increasing EHR data integration in inpatient, ambulatory, and home healthcare settings. Additionally, a growing number of patients are being monitored at home. Data are sent periodically to information systems to be reviewed and acted on by physicians, care providers, and case managers.
Given the potential benefits of the technology, MDI is a nascent strategic enterprise business and clinical imperative, says Farrell. “It’s important to start with IT objectives tied to organization goals,” she says. “Accountable care organizations and similar models will increasingly monitor patients and manage devices across community, progressively more at home. Today, far too many MDI initiatives are focused on device connectivity and not business or clinician goals. Initiatives are project or department driven, not based on an enterprise or community vision and supportive device strategy.”
Deborah Kohn, MPH, RHIA, FACHE, CPHIMS, FHIMSS, an HIM expert and principal at Dak Systems Consulting, sums up the importance of quality data for HIM professionals: “Data integrity is vital to maintaining clinically and legally sound, court-ready EHR information. HIM has a clear role and a vested interest in MDI initiatives given the profession’s advocacy for quality healthcare though quality information.”
— Leslie Feldman is a freelance healthcare writer and marketing communications consultant in Philadelphia.