Pathology Service Enhancement

---

Moving Expertise from the Laboratory to the Point of Care

Throughout the last decade Drs. Smith and McNeely have conducted research aimed at improving clinical laboratory utilisation. This research has produced a technology that has been shown to have dramatic implications for clinical laboratories, patient care and healthcare economics.
At the base of this development is the notion that if one can move the expertise of clinical pathologists from the laboratory to the point of care in an easy to use and effective manner, profound affects on laboratory utilization, healthcare economics and healthcare itself will be realised. Scientific experimentation bears this out.

Introduction
Pathology is a fundamental lynchpin of the healthcare system due to its central role in diagnosis. Over 70% of diagnoses depend on laboratory testing1. Accurate and timely diagnosis is, in turn, key to an efficient and cost-effective healthcare system. Healthcare systems remain stressed and given the critical role of pathology it is not surprising that the field operates under some significant pressures of their own:
Although laboratory medicine is vital to current medical practice, there is a more fundamental problem than those mentioned above: the expertise in laboratory medicine resides in the medical laboratory but it is physicians at the point of care that govern use. This amounts to a built in inefficiency. Research bears this out. In a review of laboratory tests ordering studies across a number of countries, researchers have converged on an estimate of 33% of laboratory tests being ordered inappropriately2.

Best practice guidelines
In terms of addressing this situation best practice guidelines represent an enormous step in the right direction. However, they have two central difficulties. First, they lend themselves to high volume testing areas that can be reduced to a few simple, widely agreed to rules. Secondly, it is very difficult to get them into routine use by ordering physicians3,4. Nevertheless, the implementation of clinical guidelines has been shown to provide significant improvement in laboratory utilisation. In British Columbia alone, the efficiencies derived from disseminating a thyroid protocol in 1992 were assessed and found to be in the order of $1million per year5.

If laboratory test ordering could be brought into greater compliance with best practices and were able to move the expertise from the clinical laboratories to the point of care, many problems would potentially be solved.
A recent study in radiology looked specifically at this approach. By providing phy-sicians with direct access to radiology specialists it was found that a 26% reduction in images ordered was possible6. Benefits lie in providing ordering advice in a fast, economically viable, convenient way that fits into the physicians workflow.

Clinical Intelligence Laboratory System
Innovative use of artificial intelligence software (Acquire®3), the Internet and wireless PDAs7 provide the wherewithal to bring the expertise of the laboratory to the point of care. In essence, to build a virtual pathologist to assist GPs with laboratory investigations. GPs, as a group, account for a large proportion of test volume and are very receptive to advice and guidance on test selection8.

The Clinical Intelligence Laboratory System (ClinLab™) is a software system that embodies best practice guidelines and the expertise of a pathologist in a test ordering and result interpretation system that runs on wireless PDAs. The guidelines and expertise are not embodied in verbose fashion, rather they are distilled to those few clinical observations that are essential to ensure that the right tests are ordered and the right interpretation arrived at. In practice this means that a laboratory investigationcan be reduced to 4–6 stylus presses per patient encounter. This makes for a very rapid and easy to use tool that can operate over existing infrastructure and conveniently fit into the clinician's pocket between encounters.

System evaluation
To investigate the viability of this approach the system was scientifically tested in Victoria, B.C., through an experiment designed to determine the impact of system usage on current practice. When comparing physicians' test ordering with and without the benefit of the system the following results were obtained through system usage:
Most importantly the beta test demonstrated that the system was able to influence the test ordering behaviour of physicians in a way that was acceptable to them and in a way that benefited both medicine and labora-tory utilization9.

Implications of these results
Each of these results has enormous implications for clinical laboratories and healthcare systems. Specifically: costs were 38% lower for the Crown.In British Columbia the projected expenditure on laboratory services was $473 million for the year 2002/200310. A 38% cost reduction would potentially lower the cost of these services for British Columbians by $180 million each year. This enormous saving could be put to immediate use in alleviating financial strain in other areas of the healthcare system.

45% fewer laboratory tests were ordered
Test volumes are rising. This puts a strain on all facets of existing facilities. Large reductions in test volume enable existing personnel to decrease the turnaround time and devote more time and resources to more complex procedures and analysis. There are enormous benefits in terms of the quality of service provided, the work environment of the personnel, the cost of hiring additional personnel and expanding or upgrading facilities to be realized through this level of demand management.
The most common outcome of outpatient laboratory testing is negative or normal results. Test cases that produced results within normal limits were separately evaluated. In these cases, physicians ordered 50% fewer tests and required 35% fewer sample collection while using ClinLab9.

25% fewer sample collections
The number of sample collections can be reduced substantially through the use of this system. The system ensures that the test ordering is complete with respect to the diagnostic dilemma in question. It utilizes an electronic connection between the laboratory and the ordering physicians such that additional tests (second or third round testing, or ordered reflexive testing) can be performed on an existing sample while it is held at the laboratory. Sample collection is a very costly and time-consuming element in the diagnostic process. Reductions in sample collections reduce costs, speed up turnaround time, speed up diagnosis and enable patients to avoid unnecessary venepuncture and inconvenience.

Far fewer referrals to specialist
Referrals to specialists can be reduced when ordering physicians are provided with the guidance to pursue the laboratory investigations in areas that they might otherwise require referring the patient to a specialist. In British Columbia the wait time between GP referral and specialist consultation is 50% longer than that deemed reasonable by BC specialists11.

Reducing referrals to specialists in this way ensures that the patient's condition does not deteriorate unnecessarily – a situation that can lead to hospitalizsation, furthering the waiting list problems. Specialists benefit as they are relieved of some of the demand on their time. The healthcare system benefits from these reductions in referrals through better allocation of resources and by decreasing the need for further recruitment of specialists.

More uniform and accurate diagnoses
Inaccurate diagnosis leads to unnecessary deterioration of patient's conditions leading to possible hospitalization and prescription of inappropriate drugs. This adds enormous cost to an already costly aspect of healthcare. In 2001 British Columbia spent $657 million on its Pharmacare program12. Given the magnitude of these expenditures it is prudent to ensure that the drugs being prescribed are for the right condition: diagnostic accuracy is key.

In a study looking at diagnostic errors in Spain it was found that major diagnostic errors were found in 3.51% of the cases investigated while minor diagnostic errors were found in 3.82% of the cases13.
If prescription drugs ordered as a result of misdiagnosis amount to 3.5% of the overall pharmaceutical expenditure then a further cost reduction of $23 million in British Columbia could potentially be realized through increasing the accuracy of diagnosis.

Closer adherence to established guidelines
Best practice is the ambition of the healthcare system. ClinLab goes beyond the areas currently covered by best practice guidelines into more complex areas and brings the test ordering and test interpretation of physicians into line with the guidance of pathologists that specialize in these areas of investigation. System evaluation found that the impact of system usage on test volume, number of sample collections and cost reduction was even greater in areas not covered by best practice guidelines.

Diagnostic turnaround time fell to one day
Reductions in the time taken to reach a diagnosis have direct implications for the care of the patient. Lengthy diagnostic investigations can lead to deterioration of a patient’s condition such that they require hospitalization. Reductions in the time taken to reach a diagnosis can therefore have a direct impact on hospital waiting lists as well as the well being of patients.

Conclusion
The clinical laboratory sector is currently at a crossroad. A crossroad that will see the information service side of this sector increase dramatically as test volumes increase and funding for healthcare is strained. In the face of this the clinical laboratories need to exert a proactive influence over laboratory utilization while providing increased information services.

These pressures on the laboratory sector coupled with the laboratory-GP-patient information channel that provide the opportunity for systems like the Clinical Intelligence Laboratory System to emerge as mechanisms for improving laboratory utilization, improving medical practice and increasing the level and quality of information being provided to GPs and their patients.

References
1. Speech by Rt Hon John Hutton MP, Minister of State (Health). Government vision for Pathology services. Speech to the Royal College of Pathologists, 22 September 2004. http://www.dh.gov.uk/NewsHome/Speeches/SpeechesList/SpeechesArticle/fs/en?CONTENT_ID=4089887&chk=eD
/aTS

2. van Walraven, C, Naylor, CD. Do we know what inappropriate laboratory utilization is? A systematic review of laboratory clinical audits (abstract). JAMA, 1998; 280: 550-558.
http://jama.amaassn.org/cgi/content/full/280/6/550

3. McNeely MD, & Smith BJ. An interactive expert system for the ordering and interpretation of laboratory tests to enhance diagnosis and control utilization. Canadian Medical Informatics, 1995 Vol. 2, No. 3, pp. 16-19. http://www.aiinc.ca/applications/las/medinfo.html

4. McAlister FA, Koon KT, Lewanczuk RZ, Wells G, Montague TJ. Contemporary practice patterns in the management of newly diagnosed hypertension. CMAJ 1997; 157: 23-30.

5. McNeely M. The Canadian experience. Seminar on Laboratory Utilization. WASP XVI World Congress, Vancouver 1993.

6. Litt, Andrew W., Ryan, Donald R. The Specialist as Gatekeeper. Decisions in Imaging Economics, Spring 2001. http://www.imagingeconomics.com/library/200103-09.asp

7. Acquired Intelligence Inc/Clinical Intelligence Ltd., Press Release. Microsoft Provides Acquired Intelligence Inc With Their 'Empower Program For Independent Software Vendors (ISVs)' To Enable Pocket PC Phone Deployment Of Our Clinical Laboratory Application. Victoria, B.C., CANADA August 12, 2004.

8. Jones, R., O'Connor, J. Information management and informatics: need for a modern pathology service. The Annals of Clinical Biochemistry, 2004, 41, pp. 183-191.

9. Smith BJ, & McNeely M. The Influence of an Expert System for Test Ordering and Interpretation on Laboratory Investigations. Clinical Chemistry, 45:8, 1168-1175, (1999). http://www.clinchem.org/
cgi/content/abstract/45/8/1168

10. Bayne, L. BC Laboratory Services Review. July 2003. http://www.plco.ca/uploads/publications/1.pdf

11.BCMA. Specialty Care In BC. A System in Distress. June 2004 http://www.bcma.org/public/news_publications/publications/policy_papers/SpecialityCare/SpecCareINTRO.pdf
12. Pharmacare Trends 2002. Pharmacare. BC Ministry of Health Services. http://www.healthservices.gov.bc.ca/pharme/outgoing/PcarTrends 2002.doc

13. Bombí J.A.; Ramírez J.; Solé M.; Grau J.M.; Chabas E.; Astudillo E.; Nicolás J.M.; Balasch J. Clinical and Autopsy Correlation Evaluated in a University Hospital in Spain (1991–2000). Pathology Research and Practice, February 2003, Vol. 199, No. 1, pp 9-14(6). http://www.ingenta.com/isis/searching/Expand/ingenta?pub=infobike://uban/571/2003/00000199/00000001/art00346
Brian A. Schaefer, D. Phil is the chairman of Clinical Intelligence Corp in Canada and the chairman of Clinical Intelligence Ltd in England and Wales.