An Historical Analysis of Pathology Ordering by General Practitioners between April 1998 and March 2001 from the Bettering the Evaluation and Care of Health (BEACH) Program (2002)
An Historical Analysis of Pathology Ordering by General Practitioners between April 1998 and March 2001 from the Bettering the Evaluation and Care of Health (BEACH) Program (2002)
Description
This project sought to analyse changes in pathology ordering patterns in general practice between 1998 and 2001 and the possible causal relationships between any identified changes and other factors by undertaking an analysis of the Bettering the Evaluation and Care of Health (BEACH) data between April 1998 and March 2001.
to investigate changes in pathology ordering patterns in general practice between April 1998-99 and March 2000-01, and the possible relationships between any identified changes and other factors
Specifically:
to investigate changes in pathology ordering patterns over three years
to investigate the extent to which characteristics of general practitioners (GPs) have changed over the three years of the program
to investigate the relationships between:
changes in pathology ordering and in GP characteristics (particularly GP age)
pathology ordering and morbidity under management
pathology ordering and length of consultation
pathology ordering and prescribing behaviour, imaging ordered, therapeutic procedures undertaken and clinical treatments provided
to investigate the factors that together significantly contribute to pathology ordering behaviour.
These aims and objectives were achieved by this project.
Findings
There has been a significant increase in pathology test order rates by GPs between 1998-99 and 2000-01.
There has only been a small increase in the proportion of problems for which pathology is ordered, but there has been a very strong move away from ordering a single test per problem to ordering three or more tests per episode.
The major increase was for requests for chemical pathology (+22.6%), with other areas showing smaller increases; Immunology (+20.1%), haematology (+13.5%), tissue pathology (+10.7%), microbiology (+5.5%), cytogenetics (+0.5%), infertility/pregnancy tests (+1.5%). There was a decrease in the number of simple basic tests (-11%) and cytopathology (-1.2%).
The overall pathology order rate increased by approximately 22%.
There was an inverse relationship between pathology test ordering and pharmaceutical prescribing, and a positive relationship between pathology test ordering and imaging test orders, but their impact on overall rates of other clinical activities was minimal.
The increases are reflected in pathology ordered for all purposes, particularly for diagnostic and preventative care and other pathology, and to a lesser extent in monitoring pathology.
The proven relationship between consultation length and pathology ordering did not have any impact on the changes in total pathology tests ordered over time.
While order rates are significantly related to GP characteristics, there has been no change in the characteristics of GPs so this relationship cannot be the cause of increased pathology test orders.
Some of the increase can be explained by increased management rates of a few common conditions, however, the majority of the measured increase cannot be explained by factors measured in this study.
External influences such as changes in the Medicare Benefits Schedule (MBS), system changes such as increased computerisation, and possibly increased fear of litigation must be considered as possible influences on pathology order rates of GPs over the period of this study.
Higher ordering rates were associated with:
larger practices
rural and remote practices
higher proportions of working-age adult patients (aged 15-64 years)
higher management rates of problems associated with the blood and blood-forming organs, the endocrine and metabolic systems, the circulatory system, pregnancy and family planning and urinogenital problems
high rates of Medicare Level C (long) consultations.
Factors which may have been influential in altering the patterns of ordering include changes to guidelines which promote more intensive monitoring of patients, a move to computerised ordering, medical indemnity concerns and increasing emphasis on managing chronic diseases.
Key Project Learnings
The increases in the number and costs of pathology tests as recorded by the Health Insurance Commission (HIC) data did not necessarily reflect the true ordering patterns of GPs.
Independent of the effect of variables associated with higher ordering rates, pathology order rates increased with time.
A significant increase in pathology test order rates by GPs could not be completely explained by factors measured in this study.
Follow on Initiatives and Projects
A second stage of this project was undertaken as Evidence-Practice Gap in GP Pathology Test Ordering: A Comparison of BEACH Pathology Data and Recommended Testing (2009).
Evidence-Practice Gap in GP Pathology Test Ordering: A Comparison of BEACH Pathology Data and Recommended Testing (2009)
Description
This study sought to identify GP pathology ordering patterns from Bettering the Evaluation And Care of Health (BEACH) Program data 2000-2007 for high cost, high use and inappropriate items to enable analysis of the patterns against published best-practice guidelines. It also sought to examine general practice (GP) pathology ordering patterns in the BEACH data for problems managed in general practice:
which were a National Health Priority Area
if pathology ordering was common in the management of the condition
if the pathology ordering behaviour of GPs had changed for the management of the problem between 2000-2002 and 2006-2008 (the duration of the study).
On this basis, Type 2 diabetes, hypertension, lipid disorders, weakness/tiredness, ‘health checks’ and overweight/obesity were selected for investigation. ‘Health checks’ problems included check-ups recorded by GPs at encounters with patients aged 15 years and over, and overweight/obesity includes problems managed that were labelled by a GP as ‘obesity’ or ‘overweight’ for patients aged 18 years and over.
All of the six problems investigated in this study accounted for 12.1% of all problems managed in 2000-08, and for more than one-quarter (25.7%) of the total pathology tests/batteries recorded by GPs.
to investigate the extent to which GPs’ pathology-ordering behaviour for selected problems aligns with recommendations made in national and international guidelines for the management of these problems (data used was from a BEACH study over eight years from April 2000 to march 2008)
to identify whether changes have occurred in the pathology ordered for the selected problems over the last eight years, and whether any measured change reflects a change to be ‘more’ or ‘less’ in line with guidelines recommendations
to identify the extent to which measured changes have been the result of changes in:
the management rate of the problem and/or
the likelihood of pathology being ordered in the management of the problem and/or
explore the number of pathology tests/batteries of tests being ordered.
These aims and objectives were achieved by this project.
Findings
The pathology test orders increased significantly for all six problems over the period of this study. These increases were independent of changes in the management rate of the problem and accounted for about 31% of the national increase in pathology tests ordered for all problems from 2000-2002 and 2006-2008.
Increases in the pathology order rates for Type 2 diabetes and hypertension were due to an increase in the likelihood of at least one test being ordered, and an increase in the number of tests ordered per tested contact.
Only the likelihood of testing increased for weakness/tiredness and overweight/obesity.
Only the number of tests ordered per tested contact increased for lipid disorders/health checks.
GP ordering behaviour aligned well with guideline recommendations for lipid disorders (75.5% of pathology tests/batteries supported), weakness/tiredness (71.7%), Type 2 diabetes (72%) and hypertension (65%). The level of support for lipid disorders and hypertension is likely to be over-estimated as tests were primarily recommended for initial assessment of newly diagnosed cases whereas the majority of pathology ordered for these problems was for ongoing management.
Pathology tests/batteries ordered by GPs for ‘health checks’ and overweight/obesity did not align well with recommended testing. Only 24.3% of pathology tests/batteries recorded for ‘health checks’ and 50.9% for overweight/obesity were recommended in the guidelines.
The number of tests ordered by GPs per tested problem increased over the period of this study and potentially further compounded the issue of deciphering true positive results (real change) from false positive results. For example, full blood counts (FBCs) were one of the most frequently ordered tests but it was only recommended in the management of weakness/tiredness and in the initial investigation of hypertension. Therefore, the results suggest GPs may have been opportunistically/routinely-ordering FBCs when ordering blood tests.
The guidelines reviewed were morbidity-based and could not provide guidance that is applicable for all patients, especially those with multiple chronic conditions.
Recommendations
Providing guidance on the variance of results in long term monitoring (coefficient of variance) and likelihood of abnormal test results when multiple tests are ordered may improve the interpretation and appropriate ordering of pathology tests in primary care.
The limitations of this study meant the level of support could not be determined for 10-24% of pathology tests ordered for each morbidity. Either tests were recommended for a specific clinical situation that could not be evaluated with BEACH data, or GPs ordered batteries of tests (e.g. multibiochemical analysis) for which support could not be determined. Further research is needed to determine whether the use of these tests is supported.
Guidelines and guidance regarding pathology tests could be improved by:
providing adequate advice on the pathology tests required in the ongoing management of each condition (e.g. recommendations regarding monitoring long term medication) including detail on the frequency and duration for which testing is required
providing advice on the pretest probability of disease, particularly when recommending investigation of possible causes of secondary disease
informing GPs of the likelihood of intra-individual variation when monitoring long-term conditions. Using medical record software to provide graphical presentation of results of repeated pathology tests with markers to indicate the coefficient of variation may be useful
educating GPs on the likelihood of false positives when ordering multiple pathology tests, particularly in the context of low pretest probability of disease
standardising terminology used to refer to pathology testing to help GPs locate information regarding pathology testing within guidelines.
The clinical indications for ordering FBCs, thyroid function tests (TFTs), multibiochemical analysis and liver function tests(LFTs) in the long-term monitoring of chronic conditions needs clarification. Further research or review of literature to determine the pretest probability of underlying disease may be useful in developing guidance on the use of these tests.
Ensuring GPs can access results of previously ordered pathology tests (regardless of who ordered the test), and that results are easily accessible within the electronic health record may decrease the rate of repeated pathology testing.
The length of guidelines was perhaps the biggest barrier to GPs using them, particularly as a quick reference point to locate information about best practice for pathology ordering, and when they are not applicable in the clinical context of multiple morbidities for one patient. Therefore, developing other avenues to provide guidance to GPs about pathology ordering may be useful such as short problem-orientated statements of recommended pathology tests relevant to the stage of management (initial diagnosis or longer-term management).
Advice on testing in long-term management needs to include information on expected intra-individual variation (biological and analytical), interval to retest and duration for which monitoring is needed. Information on the likelihood of false positive results when ordering multiple pathology tests should also be provided.
Where there is no evidence available, problem-based consensus statements should be developed with involvement of practicing GPs. These new guidance statements could be incorporated into decision support systems within electronic health records, linked at the point of the decision to order pathology tests for that problem and at the point of receipt of results.
Key Project Learnings
Locating relevant Australian guidelines was not a straightforward process. There was no central listing of the available evidence-based guidelines, and the organisation creating the guideline/guidance document varies depending on the morbidity. The National Institute of Clinical Studies, which is part of the National Health and Medical Research Council (NHMRC), was developing a national clinical practice guidelines portal at the time of the report.
Guidelines are usually not designed for GPs. They are often very long documents (often 200+ pages long) and it is unrealistic to expect GPs to read long documents for all the morbidity types they management.
Information regarding recommended pathology testing was often difficult to locate in guidelines as there was often no specific section that addressed investigations to be done. There was also mixed terminology used within the guidelines to refer to testing such as ‘diagnostic testing’, ‘laboratory investigations’, as well as in the specific test name or the disease to be tested for.
Recommendations regarding pathology testing in the long-term management of conditions were often not provided in guidelines.
The variation of test results was often not discussed in guidelines or other sources of GP guidance, with most guidelines providing a ‘target level’ without providing further detail.
Follow on Initiatives and Projects
The development of the NHMRC clinical guidelines around obesity.
Areas for Future Consideration
Requester education and decision support guidance statements, especially for the use of FBC, TFTs, multibiochemical analysis and LFTs in monitoring chronic disease.
Liaise with NHMRC/guideline developers regarding:
adding an ‘investigations’ section in guidelines
standardising pathology terminology
including pathology representation on guideline development working groups.
