May 2011
Feature Story

Erlo Roth, MD, MBA
Paula Garrott, Ed.M, MLS

A few decades ago physicians ordered whatever clinical laboratory tests they wished, laboratories obligingly performed them, and health insurers paid for them. If a test result did not fit the physician’s diagnosis it was often dismissed as a laboratory error because the clinical diagnosis reigned supreme and unchallenged.

Then some pathologists became concerned about proper lab utilization. Soon health care insurers and Medicare saw this opportunity for cost saving, and they developed algorithms for denying payment for tests that didn’t match the appropriate disease codes.

On the other hand, for clinicopathologic conferences, in which it was formerly expected that the discussant would make the diagnosis before laboratory data were given, increasing amounts of test results began to be provided to the discussant in advance, which was recognition of their importance in arriving at the correct diagnosis. Labs also began to provide interpretive reports, though they were somewhat concerned that clinicians might consider it an invasion of their turf. As interpretive comments became more accepted and clinicians even began to ask for them, pathologists provided increasingly valuable diagnostic comments. In anatomic pathology, pathologists have provided specific diagnoses since inception, and the Bethesda System has long expected the cytopathologist to include recommendations for follow­up of abnormalities.

Unfortunately, in clinical pathology, many pathologists and other laboratorians created interpretations without the appropriate knowledge. A study by Lim, et al¹ showed that only about half of the interpretive comments were correct, potentially leading to misdiagnoses. This indicates a need for a larger number of laboratory professionals knowledgeable in the clinical correlations of laboratory tests.

Automatic computer-generated comments that are not patient-specific are often insufficient to guide the clinician. That is in part why health insurance companies deny reimbursement for these services.

St. Peter, et al² reported a poll in which 38 percent of specialists felt that primary care physicians were unable to keep up with the rapid growth of new clinical information. Clinicians are gradually sacrificing their diagnostic knowledge to keep up to date their disease management knowledge. The problem may become more acute if and when the medical home concept of the new health care act is implemented. This knowledge gap, however, is not limited to primary care physicians.

At Massachusetts General Hospital a physician overlooked a case of von Willebrand disease in a coagulation profile, and an obstetrician misinterpreted a low protein S in a pregnant woman who then had an unnecessary abortion. This problem led Laposata, et al³ to initiate a program of patient-specific clinical pathologist consultations. A followup survey of clinicians revealed that in 70 to 80 percent of cases these reports saved them time and improved their diagnostic accuracy.

The importance of the laboratory’s diagnostic support of physicians is reinforced by evidence that 60 to 70 percent of the decisions on admissions, medications, and discharges are based on laboratory data.⁴ With the rapid development of genomics, proteomics, and pharmacogenomics, the clinicians’ diagnostic knowledge gap is likely to grow, and with it the demand for support from the departments of pathology.

One of the hottest topics at the September 2010 Association for Pathology Informatics meeting in Boston was the concept of combining information from clinical pathology, anatomic pathology, and radiology into a single, integrated diagnostic report. Michael Laposata, MD, PhD, and associates at Vanderbilt University are implementing this form of reporting; it resembles the conclusions at clinicopathologic conferences. The coagulation and hematopathology areas are now issuing integrated diagnostic reports, which have resulted in a one-day reduction in patient length of stay.⁵

Facilitating such reports will be the ever more powerful imaging techniques, the rapid growth of pathology informatics, including expert systems, and the increasing use of image analysis in cytopathology, immunohistochemistry, and hematology. Therefore, the future practice of pathology and radiology is likely to consist of a larger network of experts linked to diagnostic centers by telepathology, teleradiology, and other data transmission systems. These centers will be staffed by generalist pathologists and radiologists who know enough about the various subspecialties and clinical medicine to be able to assemble all the pieces of the “diagnostic Lego kit” into a coherent and comprehensive diagnostic report. In keeping with the “omics” trend of genomics, proteomics, and pharmacogenomics, this emerging diagnostic medicine “specialty” could well be called “diagnomics.” It has the potential to reduce errors in medical diagnoses and reduce the costs associated with them.

For this to become a reality, there are still obstacles to overcome. One is graduating medical students’ declining interest in clinical pathology, driven largely by reimbursement issues. However, if the services of integrated diagnostic reports are properly remunerated, as they have been in Dr. Laposata’s experience, this trend is likely to be reversed, particularly because clinicians will perceive clinical pathologists to be as valuable to them as anatomic pathologists are to surgical specialists. And because these services are patient-specific, the longstanding challenge of getting reimbursed for a clinical pathology component will have been sidestepped.

Even so, there will not be enough clinical pathologists to meet the demand for these consultations. A solution is to get medical laboratory scientists (MLS) more involved in selecting and interpreting laboratory data in their areas of expertise. A significant number of MLS programs, such as the one directed by one of the authors, include clinical correlations in their curricula. However, when these graduates start their careers, they are often inundated with the analytical processes and not given the opportunity to use their knowledge in clinical correlation. If clinical pathologists invited these medical laboratory scientists, particularly those with graduate degrees, to form a diagnostic team, the effective shortage of clinical pathology consultants would be ameliorated and the MLSs would become eligible for salaries commensurate with their capabilities. (A good parallel are the clinical pharmacologists, who stepped out of the pharmacies to become indispensable consultants to clinicians on drug therapies.) This and the professional recognition gained in their new consultative role may go a long way toward making medical laboratory science a more attractive career to high school students.

In summary, the laboratory that passively reacts to a test order will increasingly fall short of its mission. It should instead take the diagnostic role into its own hands and allow the clinicians to focus on managing disease. This will bring greater recognition to laboratory professionals, improve the quality of health care services, and reduce the costs associated with those services.

References