Transcription and EHRs: Benefits of a Blended Approach

By Jay Cannon and Susan Lucci, RHIT, CMT, AHDI-F

Mixing transcription and templates for physician documentation offers a practical way forward in organizations transitioning to electronic records.

One of the biggest roadblocks to successful adoption of electronic health record (EHR) systems concerns physician documentation; specifically, use of point-and-click, structured templates. While hospital administrators and CIOs sometimes prefer full utilization of the EHR’s structured documentation capabilities over narrative dictation and transcription, physicians struggle with capturing the complexity of their patients’ stories within checkbox templates. Many prefer traditional narrative dictation and transcription.

HIM professionals and CIOs face a considerable challenge: how to balance physician productivity, satisfaction, and preferences with the need for structured, discrete data and meaningful EHR adoption.

For physicians, every minute counts, and template-based documentation have the unintended consequence of lowering physician productivity. In addition, templates can tempt busy physicians to simply copy and paste documentation, thwarting HIM mandates for complete and accurate reports.

Likewise, CIOs are facing a technology challenge. They must establish data-reporting infrastructures to support internal and external clinical outcomes reporting, statewide hospital reporting programs, application for HITECH funding, and preparation for broader transparency and accountability. However, those infrastructures must balance physician productivity and satisfaction against the heightened discrete data reporting requirements. It is a difficult balance to achieve and even harder to maintain.

Some organizations are approaching the challenge with a blend, integrating medical transcription as a component of the EHR. By doing so, HIM professionals can work with CIOs to balance physician satisfaction, achieve meaningful use of EHRs, and most importantly, ensure accurate clinical documentation for quality patient care. Perhaps transcription has found a solid role for the future: an integrated partnership with the EHR.

Transcription and the EHR: Reality versus Perception

Over the past decade, most EHR return-on-investment calculations have included the assumption that physicians would adopt template-based documentation, and medical transcription costs would be significantly reduced or eliminated. When return is based largely on reduced transcription costs, EHRs almost always fall short.

A survey of 2,212 providers conducted by the AC Group one year following its purchase of an EHR system found that 53 percent of physicians had gone back to dictation or handwriting and 18 percent had stopped using the EHR all together.1 HIMSS Analytics reports only 27 hospitals nationwide at the stage at which the majority of physicians use structured templates for their clinical documentation.

Provider organizations that originally hoped to entirely eliminate their transcription costs have discovered that approximately 30 percent of transcription still remains. For example, Fallon Clinic in Worcester, MA, was only able to reduce transcription costs by a third, far lower than the 75 percent reduction projected by its EHR vendor.2

A survey of HIMSS Analytics stage 6 hospitals-those almost completely automated and using paperless medical records-revealed a mix of report capture options that included speech recognition, voice recognition, and structured EHR templates (see table). Hospitals in the survey averaged 35 percent use of structured templates within the EHR, 62 percent dictation and transcription, and 4 percent voice recognition.

At Mayo Clinic in Jacksonville, FL, 70 percent of physician notes are created using dictation and transcription, estimates Andrea M. Seymour-Sonnier, director of appointment and transcription services. The use of structured templates is growing (currently 25 percent of total notes are created using structure templates), but more than 620 dictators still use the organization’s medical transcriptionist and editor services, Seymour-Sonnier says.

For these organizations and thousands like them, a blended approach to physician documentation appears to be the norm. According to Mark Anderson, CEO of the AC Group, this approach will probably be the most ideal for physician documentation needs in the foreseeable future.

Mixed Approaches Even in Tech-Savvy Hospitals

A survey of hospitals rated “stage 6” by HIMSS Analytics-those almost completely automated and using paperless medical records-revealed a mix of report capture options in use.


Bed Size

% Using Structured Templates Within the EHR

% Using Dictation and Transcription*

% Using Front-End Voice Recognition (excluding radiology)












> 1,000





> 1,000










< 100




























*With or without back-end speech recognition technology
Source: e-mail survey conducted by the Friedman Marketing Group, November 2009

The Blended Model

With the blended approach to physician documentation, different modalities are used to capture dictation based on physician preference, practice patterns, and document types. For example, structured history and physical templates populated by a physician assistant may be used in one care setting, while a dictated and transcribed narrative report may be the best documentation method for inpatient discharge summaries, encounter notes, findings, and assessments.

Physicians use templates where and when they make the most sense to them and still retain the option to dictate the details of a patient visit. The approach streamlines processes without sacrificing the unique facts that impact quality patient care.

Trending at Mayo Clinic Jacksonville also indicates that a physician’s technical expertise drives his or her documentation approach. Even within the same practice it is not uncommon to find variability, Seymour-Sonnier says.

Eve-Ellen Mandler, RHIA, MS, CCS, FAHIMA, is director of HIM and privacy at St. Clair Hospital in Pittsburgh, PA. She says a blended, hybrid approach to dictation and transcription gives HIM the best of both worlds. “Physicians get the information they want and need into the reports without changing their practice patterns while HIM reduces cost and improves efficiency,” she says.

St. Clair uses back-end speech recognition technology for physicians to dictate their reports, with a few dictation templates for “normal” reports in use. Mandler reports 90 percent physician compliance, with the physician dictations being recognized by the back-end speech recognition system and resulting in good drafts. In addition, a few physicians are using structured, point-and-click documentation templates within the EHR as part of a pilot program for in-house progress notes.

Mandler recommends the blended approach. “Structured templates may take away the individual physician’s style. Even if you set up individual templates, there are fine nuances of each case that must be captured,” she says. Finally, template-based documentation creates readability challenges for the next clinician treating the patient.

“Organizations must work closely with their medical staff to achieve consensus about the proper design and formatting for documentation templates in order for them to be useful and functionally rich,” says Rich Schaeffer, CIO at St. Clair Hospital. “An off-the-shelf template is probably not going to work in all situations. It’s a matter of working closely with the medical staff to arrive at a consensus.” From Schaeffer’s perspective, the focus should be on trying to consistently improve the physician progress note using better tools, not on replacing transcription.

The Rockwood Clinic Experience

Rockwood Clinic in Spokane, WA, also uses a blended approach of narrative transcription and structured templates. Lyn Willett, RHIT, CHP, director of HIM and privacy officer, endorses the use of dictation and transcription alongside structured templates for certain specialties and physicians.

“Implementing an EHR sounds easy, but before you leap into implementation you have to take a serious look at workflows within your organization and assess the various documentation needs and styles of your providers,” Willett says. “If you don’t take a serious look at physicians’ practices and their unique documentation needs you might be surprised by their pushback.”

While it seems like EHRs are flexible and can be adapted to meet the needs of all providers, Willett found that some specialty areas have difficulties with template-based clinical documentation. The time required for providers to become familiar and proficient with template-based documentation can easily result in lost productivity, delayed patient care, and provider dissatisfaction. “We don’t want our physicians to continually struggle with templates, see fewer patients, and lose revenue if template-based documentation is difficult for them,” she concludes.

High-profile revenue and risk areas such as cardiology and urgent care have unique needs and have historically dictated lengthy patient histories in order to ensure that all the necessary details are documented. In these areas it is important for HIM professionals to carefully weigh the benefits of EHR templates against physician productivity, patient care, transcription costs, and revenue.

Every month Willett reviews medical transcription costs for all providers. Each provider is compared and contrasted with other providers within their specialty and other specialty areas. This information is also forwarded to finance every quarter.

“This allows our administration to discuss the variations between providers and look for opportunities to assist providers through the EHR transition,” Willett says. “Our providers have done an incredibly good job of improving their use of template-based documentation without placing themselves, the organization, or our patients at risk.”

The consensus among Rockwood physicians and staff is that the EHR gets high marks for readability, accessibility, and data analysis, while the narrative dictation and transcription provide a more complete and accurate picture of care, Willett says.

Mining Narrative Reports for Discrete Data

According to the Gartner Group, today’s dictation and transcription solutions are in the final phase of a three-phase evolution.3 In this phase, physicians use voice recognition to create narrative reports with three options for document editing: physician real-time self-editing, scheduled self-editing, or editing by a medical editor using speech recognition.

Also termed “once-and-done,” third-phase solutions support physician choice and provide a new, critical capability: data mining discrete information from narrative dictations. “Phase three is where real impact for patient care takes place relative to benefits for improving patient care and addressing ARRA meaningful use reporting,” says Michael W. Davis, executive director of HIMSS Analytics.

The discrete data derived from narrative reports have been called “narradata.” Once the report is edited and finalized by the physician, narradata can be immediately available for secondary uses such as clinical decision support system input, EHR data fulfillment, and other reporting requirements. As more and more reporting requirements arise, narradata may help bridge the gap for organizations that have not yet implemented structured physician documentation templates but are still required to submit quality-related data.

Phase three offers the potential to improve report turnaround, reduce costs, and meet EHR data fulfillment requirements without forcing physicians to change their documentation habits. The faster information is captured and completed, the sooner it can be used. Instead of waiting 24 or 48 hours for a narrative report, information is available in real time and becomes infinitely more valuable. Data mining tools are now embedded into most transcription systems and workflows.

“Incremental improvements dramatically improve healthcare,” Davis says. “The ability to capture narrative speech, convert it to text, apply natural language processing technology, and then upload discrete data elements into clinical decision support systems for real-time patient care is one of these incremental improvements and the ultimate goal for these systems.”

As information is being dictated, it is also being structured, tagged, and coded. According to Anderson of the AC Group, “discrete-reportable transcription is fast becoming a requirement for successful EHRs, further promoting transcription’s role as a friend of the EHR.” One example of how HIM professionals are taking advantage of discrete-reportable transcription and narradata is in support of core measure reporting.

Narradata Support Core Measure Reporting

For WellStar Health System in Atlanta, GA, narradata derived from transcribed reports are used to support CMS Core Measure reporting for acute myocardial infarction (AMI). According to Beth Kost-Woodrow, RHIA, assistant VP and chief privacy officer, it is an automated process that electronically reads dictated reports, finds specific key words and phrases related to AMI anywhere in the transcription, and then provides an electronic listing of cases to the organization’s core measure nurses. Armed with this list of in-house AMI patients, nurses have the ability to promptly review cases on the unit while the patient is admitted to the hospital and begin completing CMS core measure abstracts.

This simple listing saves nurses time reviewing admission reports and finding patients with any type of documented AMI. The technology’s next step is to pull additional core measure data elements from the transcription.

“Core measure abstracts often contain upwards of over 60 data elements that must be manually abstracted by our team,” Kost-Woodrow says. “That’s a lot of valuable time spent simply reviewing cases and manually pulling core measure data elements from various parts of the medical record including transcribed reports.”

Eventually Kost-Woodrow hopes to replace core measures man hours with narradata from transcription and speech-generated documents, thereby gaining additional value from her existing systems. “Not only will our core measures nurses receive an action list of AMI cases every morning for their rounds, as they do now, but they’ll also see a boost in productivity as they eventually use the system to mine additional data from the narrative reports,” she says.

For Sondra Hess, CCS, CCDS, clinical documentation improvement specialist at Alvarado Hospital in San Diego, CA, the same workflow applies. “I can easily find patients with heart failure or pneumonia without wasting time doing a general search on every inpatient chart,” she says.

Leah Scalf, RN, NE-BC, is a data outcomes specialist for St. Francis Beech Grove Hospital in Indianapolis, IN. She says that in addition to finding patients with admitting diagnosis of congestive heart failure or AMI, she can identify patients with a history of these diagnoses.

“Every day the system finds two or three patients I would have missed otherwise, and since it is available online, I can easily pull up my patient action list at any time and from any location,” she says.

Beyond core measure reporting, third-generation transcription systems also give HIM professionals a distinct advantage by providing a flexible starting point for the EHR and perhaps even supporting better system adoption. Templates are implemented incrementally and partially populated by other systems, ancillary clinicians, and when appropriate, narrative reports.

The Health Story Architecture and Its Impact on the EHR

The need to capture the unique details of a patient’s story is so important that an alliance of healthcare vendors, providers, and associations joined three years ago to address the issue. The Health Story Project is producing data standards that will allow transcribed narrative notes to be electronically transferred into EHR systems.

Working with Health Level Seven and other organizations, the Health Story Project develops HL7 Clinical Document Architecture implementation guides for common types of documents, brings them through the HL7 ballot process, and promotes their adoption within the industry. Four technical implementation guides are now available as HL7 standards and have also been adopted by the Department of Health and Human Services.

AHIMA, the Association for Healthcare Documentation Integrity, and the Medical Transcription Industry Association are all Health Story members.

According to Davis of HIMSS Analytics, the Health Story Project brings standardization of transcribed documents to the EHR, facilitating easier exchange of reports. “Doctors know exactly where to find the information they need if reports are in Health Story-compliant format,” he says. In addition, using the standards ensures that data can be electronically abstracted from reports for the creation of patient summaries-an essential ingredient to compliance with meaningful use criteria under ARRA’s HITECH Act.

For HIM professionals, being “Health Story compliant” ensures that a transcription vendor can feed EHR templates and systems electronically while also being fully minable and research-ready. Health Story standards provide HIM with the attributes of a patient’s unique set of circumstances as well as structured detail for coding optimization, as will be needed with ICD-10.

Flexible Starting Point, Manageable Pace

Having moved from traditional dictation and transcription to editor-based models and subsequently the once-and-done approach, third-generation systems blended with structured templates have three distinct advantages. They do not disrupt physician workflow, facilitate better physician adoption, and achieve faster meaningful use of the EHR.

This blended approach was particularly beneficial at Rockwood Clinic as physicians were slowly introduced to template-based reporting instead of being forced into them via the “big bang” approach. Physicians were given a choice of input technologies, including dictation and transcription.

“We are very sensitive to the amount of change our physicians can effectively manage,” Willett says. “Moving all 220 physicians from narrative transcription to template-based documentation all at once would have been overload for our physicians and our organization.”

As a result, the organization has eased the transition for many providers by allowing a longer learning curve for those struggling with the transition from paper to electronic records. The focus is on the use of information to improve quality, cost, and performance rather than the technology itself.

This step-by-step implementation process combined with a blended approach to physician documentation supports productivity and satisfaction. According to a recent KLAS survey of community hospital executives, “ease of physician adoption is the most important criteria for selecting health IT systems.”4 The continued use of narrative documentation options make other EHRs more user-friendly and successful.

It is important to note that while EHRs and the progressive use of structured templates is the right direction for healthcare, there are points where templates fail the physician and result in unintended consequences. Many physicians agree that cookie-cutter templates and patient notes with no uniqueness are challenges to creating a complete health story.

As physician needs, documentation requirements, and EHR use incentives combine, organizations may seek to leverage dictation and transcription in conjunction with their EHRs.


  1. Anderson, Mark R. “Digital Medical Office of the Future.” October 14, 2009. Available online at
  2. Baldwin, Gary. “Measuring ROI.” Health Data Management, May 2009. Available online at
  3. Hieb, Barry R. “The Evolving Model of Clinical Dictation and Transcription.” Gartner Industry Research. June 13, 2007.
  4. “Disruption in Community HIS Purchases: It’s All About Physician Adoption.” KLAS research report, September 2009.

Jay Cannon ( is president and COO and Susan Lucci ( is vice president of field operations at Webmedx.

Article citation:
Cannon, Jay; Lucci, Susan. "Transcription and EHRs: Benefits of a Blended Approach" Journal of AHIMA 81, no.2 (February 2010): 36-40.