Clinicians are able to insert their thoughts and assessments into the documentation,
allowing subsequent caregivers to understand and coordinate care more effectively.
Poor-quality notes preceded EHRs, but the problem was less severe in a paper record. Each note was only available to a few readers in one location at a time. In the electronic world, poorly written notes are both widely read and widely available. Yet, the etiquette and the professional peer expectations of writing in an increasingly transparent clinical record are still developing.
ARRA meaningful-use incentives may amplify this issue, as patients will be expected to gain easy and timely access to their records, and important information will be expected to follow the patient across care settings. Amid all the data there will be an increasing need for clear, concise information.
As patients transition across multiple settings for their healthcare, thoughtful coordination of care will not occur automatically; it will take both the intent to communicate and some effort. Our challenge in health-information technology (HIT) is to ease the way time-pressed clinicians communicate with each other in a world where face-to-face contact between clinicians is less common.
The use of voice-recognition technology is helping to better meet this need. Through word choice and syntax, spoken and written language conveys both explicit and implicit information in a way that drop-down choices simply cannot. Structured documentation is a critical component of the U.S. HIT strategy and should be a core requirement of data-driven health improvement. It should be balanced, however, by nuanced-free text that explains what the data means in the context of the patient's care plan and personal life.
Coordinating more effectively
Using voice recognition, clinicians are able to insert their thoughts, assessments and concerns into the documentation in real time, allowing subsequent caregivers to understand and coordinate care more effectively.
Early in the rollout of our office EHR, we began giving providers a laptop with Nuance's Dragon Medical loaded on it. We were surprised how this reduced one of the largest root causes of provider anxiety: “I can't type.” By incorporating Dragon Medical into the training, providers were ready to use it along with the mouse and keyboard.
The program required a strong PC processor and at least two gigabytes of memory. In order to move the program into the exam rooms, we needed to create a roaming profile via an HTML server. Once this was complete, however, the provider could use the program on any exam-room computer with no loss of accuracy. By adding local vocabulary and provider names, the recognition accuracy for providers was increased.
Importantly, voice recognition was treated as a component of the EHR strategy rather than as a separate initiative. At go-live for the EHR, the information systems training staff coached the new users on Dragon Medical in the context of going live with the EHR. With this approach, more than 90 percent of long-term acceptances by providers was achieved. The remaining group split between those who stayed with dictation out of personal preference, and those who typed so fast that they did not need voice recognition.
In my practice, I use voice recognition in the exam room with patients. When it comes time to review the plan, I am able to speak in clear language as they watch the words appear on the monitor. When they are ready to leave, our staff prints a copy of the note for them. This allows them to both review and share the information when they get home.
More attention to patients
In addition to the improved transparency of this work flow, there is an efficiency to dictating in front of patients, as the provider only has to articulate the plan once, instead of having to repeat the story a second time to the transcriptionists. The clinician also validates to the patient that they have listened attentively, while the patients have the opportunity to correct any information that the clinician may have misheard. All of this serves to reinforce the patient's perceived value of the encounter.
Nationally, most of the quality-control measures for provider documentation have focused on either billing or data-driven measures. Review of the record for its communication quality has largely been neglected. Discharge documentation created at the end of a hospitalization may be audited for its billing compliance, and the provider may be held accountable for timely completion; but there is too often no quality assurance to review how accurately and completely it explains the hospitalization, or whether it provided a clear handoff to the subsequent care team. This qualitative assessment may be more subjective than quantitative measures, but clinicians can quickly distinguish between an effective and ineffective note.
We have recognized we need to deliberately prioritize communication in our care processes. This is ultimately a professional issue that must be led by clinicians who are willing to set and uphold standards. With the expansion of hospitalists and the pace of office care, synchronous voice communication is increasingly difficult. Thus, asynchronous communication by electronic messages or within clinical notes becomes increasingly critical.
While the deployment and wide acceptance of voice recognition has reduced our transcription costs by more than 90 percent, it would be a mistake to view this as a business-driven initiative. By telling the patient's story in clear words that have meaning to the whole care team, including the patient, both our patients and their specific problems can receive better care.
R. Hal Baker, M.D., is vice president and CIO of Wellspan Health, an 8,400-employee integrated delivery system in south central Pennsylvania.
For more information on Nuance solutions:
www.rsleads.com/005ht-204
May 2010
