January 17, 2011
By Alice Shepherd
For The Record
Vol. 23 No. 1 P. 10
By using an EHR template to more effectively gather family histories, a California health system has been able to send patients on a path toward better health.
Family health history information is an effective and accurate tool for assessing an individual’s risk of developing various diseases. Healthcare providers who collect and evaluate this data in clinical care settings can identify high-risk patients and encourage them to seek genetic consultation, take precautionary measures such as early and frequent screening, and make behavioral and lifestyle changes. In particular, family health history assessment has proven to be an excellent genetic screening test in cancer cases, so much so that it is considered the gold standard for initial assessments.
To help promote this important health tool, several federal agencies and professional organizations have developed comprehensive online programs that allow patients to collect, record, and organize their family health history and share it with doctors. EHRs are viewed as an important vehicle for gathering family data that can be used to improve patient care and lower costs. One healthcare system has chosen the EHR route to get a head start on identifying at-risk patients.
The VA Takes the Lead
The VA Greater Los Angeles Healthcare System (VA GLA) has developed and implemented an education program designed to improve familial risk assessment for two common hereditary cancer syndromes: hereditary breast-ovarian cancer and Lynch syndrome. The goal is to increase family history documentation and appropriate referrals for genetic consultation.
VA GLA providers collect the information using a template designed specifically for the VA’s VistA EHR. Developed by Maren Scheuner, MD, MPH, director of the Health Services Genomics Program and clinical genetic services at VA GLA, and her colleagues, the program has resulted in better cancer family history documentation, improved recognition of patients at risk for hereditary cancer, and increased referrals for genetic consultation and testing.
Seven primary care clinicians, who see about 4,000 patients per year at the VA GLA’s two women’s clinics, are participating in the three-year project. Its components are informational (a website and a continuing medical education-approved lecture series for clinicians), clinical (patient and provider information sheets and a family history reminder within the EHR template), and behavioral (practice feedback reports regarding clinicians’use of the EHR family history reminder).
During the first five months of the project, clinicians captured 568 cancer family histories, including detailed data that would not have been previously documented but is necessary to recognize high-risk patterns. As a result, 47 patients have already been referred to the VA GLA’s genetics consultation service for risk assessment and possible genetic testing—referrals that quite possibly would not otherwise have been made.
Of the various program components, the EHR cancer family history reminder and template appear to be having the greatest impact on encouraging clinicians to systematically collect family history information. The template’s opening page asks clinicians to click “yes,” “no,” or “don’t know” to seven stem questions. One question is “Were any first-degree relatives (parents, siblings, children) affected with cancer?” If yes is clicked, another window opens asking the clinician to select the type of relative, the type of cancer, and the age of onset.
When the patient has little or no family history of cancer, the clinician can quickly click through the template. Even when there are yes answers, it takes only a couple minutes to complete all the information.
“My research team conducted interviews with the providers and learned that in most cases it did not interrupt workflow,” says Scheuner. “Sometimes providers could not complete the task on the day assigned because of other pressing issues.”
As providers are clicking through the template, the program automatically generates text for the progress note (eg, “Mother history breast cancer, age at onset < 50 years”), the type of information providers without a template would have to type manually. On the last page of the template, clinicians are asked to select one of three options:
• Request genetic consultation for cancer. (Order screen will open when you click on the “Finish” button below.)
• Patient declines genetic consultation for cancer.
• Genetic consultation for cancer not indicated.
“Even when a genetic consultation is not indicated because a patient does not meet the criteria for hereditary risk, clinicians can use the family history information they gather to help guide their recommendations for screening,” says Scheuner. “For instance, if a close relative had colon cancer and there is no other family history, the provider might advise the patient to begin colonoscopy at age 40 or 10 years before the youngest diagnosis in the family, whichever comes first.”
Since VA providers have used the VistA EHR since the 1980s, the participating clinicians were familiar with similar templates, along with the electronic reminders to complete them. No technology training was necessary. However, prior to introducing the clinical tools at the women’s clinics, Scheuner’s team pilot-tested them with other VA GLA primary care providers. Feedback was used to develop the most effective design for the template and information sheets.
The patients who participate in the program are not specifically selected but are individuals who have regularly scheduled appointments at the participating clinics. During check-in, front office staff provide them with a family history questionnaire they can complete while waiting to see the clinician. Seeing the reminder to complete family history information in the “active reminder” list in the patient’s medical record, the clinician then reviews the questionnaire completed by the patient and verifies the information while clicking off answers in the template. When patients do not have time to complete the questionnaire, the clinician asks the questions during the appointment. On occasion, patients are preoccupied with their complaints and decline to participate, in which case they may be asked again at their next visit.
While an assessment of patients’ reaction to the program is not part of the study, the patient questionnaire was piloted with veterans through the VA’s Patient Education Resource Center to ensure appropriate wording and language level. “The questionnaire was readily accepted by the veterans in the pilot who stated that they were accustomed to sharing family history information with their providers,” says Scheuner.
To introduce clinicians to the program, over a period of seven months Scheuner conducted a series of monthly lunchtime lectures at the two participating clinics. Case based and interactive, her presentations focused on the importance of recognizing patients with an increased risk of developing certain cancers. Participating providers were also personally contacted to ensure a good understanding.
Further education was offered in the form of a SharePoint website. “The site’s development was a team effort with our advisory board,” says Scheuner. “They provided input regarding the basic principles, concepts, and terminology that should be taught to providers. The website includes information from trusted sources such as the CDC [Centers for Disease Control and Prevention] and other organizations with relevant information about cancer genetics. It has a discussion board and links to research articles. However, providers are not using the site very much. Feedback indicates that they prefer to rely on the core information presented at the noon lectures.”
The initiative includes multiple modes of quantitative and qualitative evaluation methods to assess changes in clinician behavior. Quarterly behavioral practice feedback reports show how many patients each provider saw for whom a reminder to document family history was due, how often the clinicians completed the task, and whether they made a referral for genetic consultation.
“Interviews with providers have told us that some really like the practice feedback reports and learn how they compare to their colleagues in the clinic,” says Scheuner. “Others don’t find the reports particularly useful. It’s a matter of personal preference. However, if family history documentation were to become a performance measure in the future, these reports could provide valuable feedback on clinicians’ performance on specific benchmarks vis-à-vis their peers.”
Each month, the EHR generates reports based on variables created in the database when a template is completed. Scheuner reviews the reports to assess a particular patient’s family history and whether a referral was made for genetic consultation.
Another source of data is a random sampling of patient charts from the participating clinics. Whether or not templates were used, Scheuner’s team reviews patient characteristics and checks whether family history was documented and which family history elements were included.
Rich qualitative data come from provider interviews. “Our anthropologist and genetic counselor interviewed all the providers enrolled in the study and asked them what they liked and didn’t like about the different interventions: patient-administered family history questionnaire, template, genetic consult service, monthly didactic lectures, website,” says Scheuner. “We plan to assess where there is variability and consensus.”
When the program was initiated, clinicians who enrolled in the study were asked to complete a 35-item needs assessment questionnaire with multiple choice questions covering knowledge and attitudes concerning genetic concepts and principles. “We have that baseline data, but we won’t have follow-up data until the program is completed,” says Scheuner. “We’re only six months into our implementation.”
Outcomes Are Promising
When Scheuner’s team examined a random sample—10% of all the charts of patients seen in the women’s clinics during each month before and after implementation—the results seemed disappointing. Thirty-two percent of the charts still showed no documentation of family history, and 18% had only a brief notation of “family history of…” that failed to specify the type of relative who had cancer. The other half of the charts, however, did provide some detail about family history.
In reviewing those charts with more specific documentation of family history, the program’s success was evident. After implementation of the EHR reminder and template, much more detailed family history was documented. Sixty-nine percent of the charts recorded a family history for second-degree relatives and 54% for third-degree relatives postimplementation compared with 33% and 5% preimplementation, respectively.
Digging deeper, 68% of the charts reflected the lineage of relatives compared with only 15% before the intervention. Documentation of age at cancer diagnosis was up from 19% to 88%; Jewish ancestry (significant for certain types of cancers) was recorded in 64% of the charts (compared with zero previously); and genetic consults were indicated in 10% of the charts (compared with none before the implementation).
Most importantly, the program is succeeding in identifying at-risk patients who may not have been identified otherwise. “The template is a big part of that,” says Scheuner. “We’ve shown that providers are in fact interested in gathering and acting on family health history information when given the education and the tools to do so efficiently.”
Scheuner would like to expand the project to other primary care clinics and perhaps to oncologists at the VA GLA—where more than 80,000 patients are seen per year—to incorporate a more diverse population. A multisite study with a control group would also be useful to assess outcomes more rigorously. Another one of Scheuner’s goals is to collect behavioral health outcomes of the patients who were part of the study. She also plans to implement telemedicine, currently used at many of the VA’s clinics, to facilitate genetic consultations for patients in her region.
Later, the program could be tested at nongovernmental healthcare organizations. To encourage clinicians in the public sector to document and act on family health history information, it may require additional CPT codes to make the detailed work reimbursable. “Family history is one of the variables included in the CPT codes, but no distinction is made between a brief notation such as ‘family history of ovarian cancer and a comprehensive assessment of the family history, including all first- and second-degree relatives, lineage, and age of diagnosis,” says Scheuner. “The American College of Medical Genetics has been trying to push through a new CPT code for pedigree analysis, but it hasn’t gone very far.”
EHR vendors, too, need to step up and develop modules that facilitate the documentation of detailed family health histories. Unfortunately, Scheuner says, most EHRs fall short of providing efficient tools to systematically document such information. While some programs have a family history tab where clinicians can enter some basic data, many providers find it easier to type brief notations in the progress notes. If healthcare professionals are to effectively collect useful, comprehensive family health histories, Scheuner says they need tools that do not disrupt workflow. They also must be educated on the importance of documenting this vital information.
“Currently many providers find it too time consuming to gather and document relevant information,” says Scheuner. “Having tools that facilitate that process without interrupting the regular workflow is extremely important in clinical practice.”
— Alice Shepherd is a southern California-based business-to-business journalist specializing in healthcare topics.