SummaryPathologists at Massachusetts General Hospital provide a paragraph of patient-specific analysis as a supplement to traditional reporting of laboratory test results. The paragraph, which takes into account factors such as the patient's age, past clinical conditions, and other related test results, may suggest a diagnosis, advise the clinician regarding additional testing, advocate the need for a specialist consultation, or suggest a therapeutic option. A team that includes a pathologist and one or more medical residents prepares each analysis after reviewing options available from a software program and discussing test results. The process has reduced ordering errors by clinicians; in addition, clinicians report that the service provides information that helps in making a diagnosis, thus saving them time and shortening the diagnostic process.Moderate: The evidence consists of pre- and post-implementation comparisons of test ordering errors, along with post-implementation feedback from clinicians using the service, as reported in two surveys (one of 49 users and a second of 100 users).
Developing OrganizationsMassachusetts General Hospital
Date First Implemented1993
Problem AddressedFaced with the rapid growth in the number of laboratory tests and the amount of available medical information, time-pressed clinicians often misinterpret test results or fail to order tests that could help in making a correct diagnosis, thus potentially compromising patient care. Many laboratories use software that provides standardized commentary based on test results, but this information often has limited value because it does not account for patient-specific considerations.
- Rapid growth in medical information, creating challenges for physicians: The number of available laboratory tests has expanded significantly with the discovery of the molecular basis for many disorders, making the selection of tests and interpretation of results challenging for physicians. Regular updates to recommended guidelines and an increasing number of studies published in medical journals add to this challenge. A survey of 7,000 physicians found that 1 in 4 primary care physicians feel that the scope of care they are expected to provide is too large.1 Another study found that patients in the United States receive only 55 percent of recommended diagnostic evaluations and treatments.2
- Time pressures that add to the challenge: Physicians cite lack of time as a primary reason they fail to obtain definitive answers to their uncertainties about ordering and interpreting laboratory tests, even when the uncertainty regards a critical diagnostic or management decision.3
- Negative impact on treatment: Although research on the effect of misinterpreted laboratory tests is limited, anecdotal reports suggest these types of errors can have serious consequences. Examples from Massachusetts General Hospital include:
- A patient had an undetected coagulation factor deficiency that should have been identified in a simple evaluation of prothrombin time, leading to a catastrophic bleed after he was taken in for neurosurgery.
- A woman terminated her pregnancy after being told by her obstetrician that she had a risk of lethal thrombosis because of a low protein S value. However, protein S values typically decrease during pregnancy, a fact that the obstetrician failed to take into account.
- In separate cases, two fathers of infants with subdural hematomas were falsely accused of abusing their children. In both cases, the hematomas had been caused by undiagnosed von Willebrand disease.
- Inadequacies of existing automated commentary: Many laboratories use software programs that provide a fixed commentary associated with a particular test result. However, such "canned" commentaries do not take into account the patient's age, medical history, and other patient-specific issues, and consequently, provide little value to clinicians.
Description of the Innovative ActivityPathologists at Massachusetts General Hospital provide a paragraph of patient-specific analysis as a supplement to traditional reporting of laboratory test results. The paragraph, which takes into account factors such as the patient's age, past clinical conditions, and other related test results, may suggest a diagnosis, advise the clinician regarding additional testing, advocate the need for a specialist consultation, or suggest a therapeutic option. A team that includes a pathologist and one or more medical residents prepares each analysis after reviewing options available from a software program and discussing test results. The service works as follows:
- Scope of service: The service provides interpretations for Massachusetts General Hospital and for more than 50 other hospitals that send specimens to Massachusetts General. The service began with interpretations of coagulation results and gradually expanded to include other laboratory tests in areas such as autoimmunity, hemoglobinopathy, human immunodeficiency virus, toxicology, blood transfusion reactions, and difficult cross-matches.
- Daily case analysis: Toward the end of each day, pathologists and pathology residents pair up to perform interpretations of test results during daily signout rounds. Other clinicians, the laboratory supervisor, medical house staff, fellows, and additional residents may also observe and/or participate, especially if the case is more complex than usual.
- Augmented paragraph based on software-generated outline: The pathologist and resident review the DxAuthor®-generated outline and discuss the case. Taking into account the test results, the patient's past clinical conditions, and other related test results, they customize the outline into a patient-specific paragraph, typically four to six sentences in length. The interpretation may suggest a diagnosis, advise the clinician regarding additional testing, highlight the need for a specialist consultation, or suggest a therapeutic option. Two examples of interpretations for coagulation test results appear below:
- For a patient with a normal von Willebrand evaluation in the presence of an acute-phase response: "The prothrombin time and partial thromboplastin time are normal. The fibrinogen, which is an acute-phase reactant, is elevated at 545 mg/dL. The von Willebrand panel values fall within normal ranges for a blood group O individual such as this patient and are not suggestive of von Willebrand disease. However, von Willebrand factor is also an acute phase reactant. Therefore, the von Willebrand panel values may be elevated above their true baseline. If indicated, repeat testing can be performed at a time when the patient is not likely to be in an acute phase reaction."
- For a patient on heparin with a low antithrombin result: "The specimen submitted has an elevated partial thromboplastin time. When the sample was treated with an enzyme that degrades heparin, the partial thromboplastin time corrected into the normal range, indicating the prolongation is due to the presence of heparin in the sample. The antithrombin is slightly low. Heparin administration can cause slight decreases in antithrombin within several days, secondary to increased clearance. If hereditary antithrombin deficiency is strongly suspected, the assay may be repeated once the patient has been off heparin for at least 1 to 2 weeks. The results of the other requested studies are normal."
- Regular updates to build knowledge base: Each interpretation is automatically added to the overall database, so subsequent interpretations can be based on an increasingly large number of cases and variables. In addition, an information technology specialist incorporates new tests and medical information related to the diagnostic process as they become available.
References/Related ArticlesKratz A, Soderberg BL, Szczepiorkowski ZM, et al. The generation of narrative interpretation in laboratory medicine: a description of individual sign-out rounds. Am J Clin Pathol. 2001;116 (Suppl):S133-40. [PubMed]
Laposata ME, Laposata M, Van Cott EM, et al. Physician survey of a laboratory medicine interpretative service and evaluation of the influence of interpretations on laboratory test ordering. Arch Pathol Lab Med. 2004;128:1424-7. [PubMed]
Laposata M, Dighe A. "Pre-pre" and "post-post" analytical error: high-incidence patient safety hazards involving the clinical laboratory. Clin Chem Lab Med. 2007;45(6):712-9. [PubMed]
Contact the InnovatorMichael Laposata, MD, PhD
Professor and Executive Vice Chair of Pathology
Department of Pathology
Vanderbilt University Medical Center
1301 Medical Center Drive
Room 4605 TVC
Nashville, TN 37232
Innovator DisclosuresDr. Laposata has not indicated whether he has financial interests or business/professional affiliations relevant to the work described in this profile.
ResultsThe process has reduced ordering errors by clinicians; clinicians report that the service provides useful information that helps in making a diagnosis, thus saving them time and shortening the diagnostic process.
Moderate: The evidence consists of pre- and post-implementation comparisons of test ordering errors, along with post-implementation feedback from clinicians using the service, as reported in two surveys (one of 49 users and a second of 100 users).
- Fewer test ordering errors: The average number of errors per requisition for hypercoagulability testing decreased significantly, from 3.56 to 1.62 during the first 3 months after program implementation. This decline suggests that the interpretations have helped to educate clinicians on the appropriate test selection pattern.
- Useful information, leading to quicker diagnosis: Two surveys found that clinicians believe that the service provides information that helps in making a diagnosis, thus saving them time and shortening the diagnostic process.
- Useful information: In one survey, 98 percent of users reported that the service provided useful information. In a second survey, 78 percent of clinicians reported that the interpretations helped them in making differential diagnoses.
- Leading to time savings and quicker diagnosis: In one survey, 59 percent of users reported that the interpretations saved them time. In a second survey, 43 percent of users reported that the interpretations reduced the time to diagnosis, with most saying that the service sped up the diagnostic process by up to 6 hours, primarily by reducing the need for additional tests, procedures, medications, and blood products.
Context of the InnovationMassachusetts General Hospital is a 900-bed medical center in Boston, MA, that includes five multidisciplinary care centers (cancer, digestive disorders, heart disease, transplantation, and vascular medicine). The laboratory medicine interpretive service grew out of a growing awareness in the early 1990s of the aforementioned challenges that clinicians face in ordering the right tests and interpreting the results. Pathologists believed their expertise could provide analytical information that could enhance the diagnostic process.
Planning and Development ProcessThe service was developed and implemented gradually over several years; key steps included the following:
- Initial implementation on small scale: The service began on a small scale in 1993, with several pathologists providing interpretations for test results related to coagulation disorders. At this point, pathologists wrote all interpretations based on their previous experience. Many new interpretations were added to a database, and pathologists began referring back to this database when writing subsequent similar interpretations.
- Expansion to other areas: Based on an early favorable response from clinicians using the service, the hospital gradually expanded the service to new clinical areas. Major additions in the late 1990s included blood bank serology rounds and hemoglobinopathy evaluations.
- Addition of software: In 1996 one of the founding pathologists gave a presentation about the service at a medical conference. A conference attendee expressed interest in developing software that could provide an outline for each interpretation based on previous similar cases and evidence-based guidelines, thereby saving pathologists time and enhancing the accuracy of the interpretations. The hospital subsequently hired the attendee to develop the DxAuthor® software, which was integrated into the service in 1998.
- Training of residents: Pathology residents work full-time with an attending pathologist and start taking an active role in interpretations on the first day of the rotation.
Resources Used and Skills Needed
- Staffing: Pathologists and other laboratory personnel participate in the program as a part of their regular duties.
- Costs: Program costs consist primarily of salary and benefits for the information technology specialist.
Funding SourcesOther than slightly upgrading the audiovisual capabilities of the conference room, no funding was provided to implement this program.
Getting Started with This Innovation
- Emphasize clinical experience and knowledge: Pathologists should have the appropriate clinical experience and knowledge base to make effective interpretive comments. To be useful, interpretations must be provided by an expert who understands the tests and clinical conditions associated with the findings.
- Prepare for some skepticism: Some veteran physicians may initially be skeptical about the interpretations, believing they have analyzed test results themselves for decades and therefore do not need outside help. Program leaders can overcome such skepticism by emphasizing that the interpretations represent suggestions, and that clinicians need not feel obliged to follow any recommendations. Over time, the service's credibility should grow as more clinicians use the service and find it beneficial.
- Visit site(s) with existing interpretive service: The program entails many logistical issues that must be addressed, such as billing, choosing software, etc. As a result, potential adopters might benefit from visiting one or more other facilities that have implemented this program to learn how they have handled these issues.
Sustaining This Innovation
- Accept the uncertainty related to program impact on costs: Improved laboratory test selection and interpretation does not necessarily translate into lower costs. In fact, in some cases, testing and treatment costs may rise because clinicians may not be ordering all necessary tests and/or the tests may identify conditions that previously went undetected. These additional expenses may be outweighed by savings generated from faster, more accurate diagnoses, but the bottom-line impact on costs may be hard to determine.
- Build in updates: Create a mechanism for incorporating new medical information and findings into the system.
Additional Considerations and Lessons
- Disclosure: Dr. Laposata is a part owner and minority stockholder of American Medical Diagnostics, the company that developed and provides the DxAuthor® software program to facilitate pathologist-generated interpretations used in this innovation.
Use By Other OrganizationsSeveral other hospitals use the DxAuthor® software to provide similar services, including:
- Cedars-Sinai Medical Center, Los Angeles, CA
- Vanderbilt University Medical Center, Nashville, TN
- Geisinger Medical Center, Danville, PA
- Children's Memorial Hospital, Chicago, IL
Kratz A, Soderberg BL, Szczepiorkowski ZM, et al. The generation of narrative interpretation in laboratory medicine: a description of individual sign-out rounds. Am J Clin Pathol 2001;116 (Suppl):S133-40. [PubMed]
St Peter R, Reed MC, Kemper P, Blumenthal D. Changes in the scope of care provided by primary care physicians. N Engl J Med. 1999;341:1980-5. [PubMed]
Green M, Ciampi M, Ellis P. Residents' medical information needs in clinic: are they being met? Am J Med. 2000;109:218-23. [PubMed]
|Disclaimer: The inclusion of an innovation in the Innovations Exchange does not constitute or imply an endorsement by the U.S. Department of Health and Human Services, the Agency for Healthcare Research and Quality, or Westat of the innovation or of the submitter or developer of the innovation. Read more.|
Service Delivery Innovation Profile
Original publication: May 26, 2010.
Original publication indicates the date the profile was first posted to the Innovations Exchange.
Last updated: October 09, 2013.
Last updated indicates the date the most recent changes to the profile were posted to the Innovations Exchange.
Date verified by innovator: April 18, 2012.
Date verified by innovator indicates the most recent date the innovator provided feedback during the annual review process. The innovator is invited to review, update, and verify the profile annually.