Quality use of pathology in general practice

Aims

The objectives of the study are to:

1. Define characteristics and changes over time in patient/general practice/pathology laboratory interactions and services;
2. Identify key areas to monitor compliance with evidence-based guidelines relating to the appropriate and quality use of pathology in general practice, and identify and measure variability using statistical process control methods; and
3. Co-design and evaluate the effectiveness of Primary Health Network (PHN)-led quality improvements.

The aim of the activity is to identify variation in pathology requests in general practice in the context of best practice guidelines, to improve the quality of pathology in general practice and patient outcomes.

Design and method

This project provides an overview of pathology testing in general practice in Australia, as well as reporting variation in testing across key indicators. These indicators were selected based on a variety of factors, including:

1. Consultation with experts and stakeholders: Recommendations were sought from experts in the field (GPs, pharmacists, and pathologists) as well as stakeholders (representatives from primary health networks (PHNs)) to determine (1) which diseases, conditions, or medications are monitored and diagnosed with pathology testing, and (2) which tests are routinely used, in general practice.
2. Literature screening: an evidence scan was conducted across peer-reviewed journals, general practice publications, the Medicare Benefits Schedule (MBS) Review, and clinical guidelines from both Australia and worldwide to determine best practice for each of these indicators, including which patients should receive specific pathology tests and frequency of testing for specific patient groups. These guidelines and recommendations formed a starting point from which to assess variation within general practice pathology testing data.

Indicators with the greatest potential for clinical impact and improvement to patient outcomes were selected to be included in the study, especially those which were of concern for specific regional areas as identified by PHNs. In many cases the selection was also made based on the availability and best quality of data within our datasets.

Findings

The outcome of this project is a comprehensive description of the use of pathology testing and variation in Australian general practice for the care and monitoring of patients. Additionally, our strong relationship with stakeholders has allowed us to provide PHNs with an evidence base for their quality improvement activities. This bidirectional flow of information has facilitated the direct translation of research to general practice.

Data overview: Active patients accounted for approximately 60–65% of total recorded patients from 2008-09 to 2017-18 in the original data. The demographic distributions of active patients in our sample and the samples collected by the national survey Bettering the Evaluation and Care of Health (BEACH) from 2015-2016 were relatively similar. The most common reasons for GP practice visits that appeared in the general practice data from Victoria and the national data (BEACH study and PBS data) were also similar. Overall, the 1o most commonly reported issues or reasons for patients’ visits in 2017-2018 were for care management (i.e., care plan and review, prescription, results discussion, and referral), preventative care (i.e., immunisation, health assessment), and the most common diagnoses were respiratory tract infection (RTI), and hypertension. BEACH similarly indicated that hypertension, check-up, upper RTI and immunisation were in the 10 most frequently encountered individual issues in all years from 2006-07 to 2015-16.

HbA1c and kidney function for monitoring type 2 diabetes: The estimated prevalence and  demographic distributions of type 2 diabetes from GP data from the three Victorian PHNs (6.2%) was consistent with the report from the national health survey from the Australian Institute of Health and Welfare (AIHW) published in 2018, which estimated 1.2 million or 6% of Australian adults had diabetes in 2017-18. While the majority of patients (71%-81%) in the Victorian data had kidney function tests annually as recommended by guidelines, only 50% or less of patients received HbA1c testing as per the guidelines’ recommendation. Improvement in follow-up HbA1c testing appeared to be particularly required in those who had results outside of the recommended HbA1c level (>53mmol/mol), where the overall median time interval was 4.7 months (IQR: 3.5–6.6 months) and slightly above the recommended three months.

INR testing for patients taking warfarin: There were substantial differences across PHNs in the proportion of patients on warfarin tested for INR. While the two metropolitan PHNs had 73– 74% of patients tested for INR within the recommended timeframe (4-6 weekly), there were only 19% of patients in PHN2, which is situated in a regional area. Further investigations are required to identify the underlying causes of the disparity in INR testing between PHNs, and where appropriate, to plan improvement activities.

Kidney function testing for patients taking Non-vitamin K antagonist oral anticoagulants (NOACs): The majority of patients (70%-83%) had their kidney function tested at least once a year as advised by clinical guidelines. The socio-demographic patterns of the patients who had annual testing however varied considerably by PHN, indicating that target populations within each PHN should be identified if further improvement in the annual monitoring is required.

Ferritin for iron deficiency: While there were no noticeable differences by PHN, a greater proportion of ferritin test results returned lower results for younger females compared to other age and gender groups. We observed higher variation among practices in the rate of low ferritin tests in females compared to males. This suggests that ferritin tests may be ordered in a higher variety of situations for females. Targeted testing focusing on adolescent and younger females may ensure capturing the groups most at risk for iron deficiency.

Vitamin D testing: Our results demonstrate an increase in the proportion of patients having vitamin D testing from 8% in 2014 to 11% in late 2018. We observed lower proportions of vitamin D tests with low test results in summer months (15-20%) compared to winter months (30-35%), potentially demonstrating lower value testing during the warmer months. We also observed high variation in Vitamin D testing among practices within PHNs, suggesting that there is a lack of consensus in regard to vitamin D testing.

Thyroid function testing (TFT): Test ordering for patterns of thyroid function testing among Australian general practitioners align closely with recommended guidelines. Approximately 80% of initial thyroid function test ordering was in the form of a thyroid stimulating hormone test, which is in line with clinical guideline recommendations. There was also low variation observed among practices in the rate of tests returning with abnormal results, suggesting consistency in ordering patterns for these tests. Between 6.0% and 11.7% of TFT tests returned abnormal results depending on PHN and gender, indicating that more targeted testing may be necessary to improve the use of this panel of tests.

Vitamin B12 testing: Overall, the proportion of patients tested for vitamin B12 was 24%, and females were tested more frequently than males (29% vs 19%, respectively). However, among older people (≥ 65 years), males were more likely to be tested for vitamin B12 (39% vs 29% form females in this age group). For patients tested on repeated occasions, the median time interval between tests was 8 months for serum vitamin B12 and 8.2 months for vitamin B12 marker, both of which were shorter than the 12-month re-testing interval recommended by the most recent Medicare Benefits Schedule (MBS) Review Taskforce. Findings were similar across PHNs.

Prostate specific antigen (PSA) for prostate cancer screening: For men without symptoms of prostate cancer (PCa), the recommended testing age is between 50–69 years. Our analysis showed that the prevalence of PSA testing among patients aged 50–69 years was 45%, while for those aged 70 years and over, it was 39%. The proportion of patients tested multiple times increased with age, with 44% of those aged 70 years and over receiving repeat PSA testing. Furthermore, among patients aged 50–69 years, the testing interval following a normal PSA result was 12.4 months (vs recommended 2 years), whereas that of abnormal PSA results was 5.9 months (vs recommended 1-3 months). While PSA testing rates varied across PHNs, the frequency of testing and re-testing interval was similar.

Bowel cancer screening and colonoscopy referral: The proportion of patients aged 50 – 74 years who had an immunochemical faecal occult blood test (iFOBT) result recorded was 0.9%. Of that proportion, 67% lived in inner regional areas. Overall, males (46%) had lower screening rates than females (54%). However, the proportion of tested males increased with age. The positivity rate overall was 5.3% (107 of 2,024 patients tested). These findings are comparable to those from the most recent National Bowel Screening Program (NBCSP) monitoring report. The low bowel cancer screening rate can be attributable to GP clinical systems not being able to extract iFOBT tests issued as part of the NBCSP.

Clozapine: With the introduction of new supply arrangements for clozapine maintenance therapy, GPs no longer need to be affiliated with a hospital to prescribe clozapine as clozapine is available at community pharmacies. However, GPs are required to review patients’ white blood cell and neutrophil counts monthly before each prescription, and follow standard clozapine monitoring protocols. Our analysis showed that 66% of patients had monthly clozapine prescriptions and 79% of prescriptions followed recommended treatment guidelines. Since the monitoring of these patients still remains the responsibility of the supervising psychiatrist and secondary care activity is not captured by the GP clinical system, these findings suggest that GP data may have limited usability in the monitoring of care delivered to patients on clozapine.

Folate: Overall the proportion of patients tested for serum folate was 17%, and females had higher test rates than males (20% vs 13%, respectively). However, among older people (≥65 years), males had high rates of testing (37%) compared to females (28%). In patients tested repeatedly, the median time interval between tests was 8.1 months, which was shorter than the 12-month re- testing interval recommended by the most recent Medicare Benefits Schedule (MBS) Review Taskforce. Findings were similar across PHNs.

Benefits

The results of this study can potentially benefit a range of different stakeholders in the healthcare system:

Patients/consumers:

Patients/consumers will benefit from Primary Health Network (PHN)-led quality improvements based on this data, leading to improvements in the care process. This will result in improved quality use of pathology in general practice. Evidence to inform these improvements can be formed from our benchmark data on pathology referral practices. This will lead to a reduction in unnecessary tests, encourage more targeted testing for those for whom it is most necessary (e.g., those at risk for certain conditions or complications), and thus allow for better use of pathology services to improve patients’ treatment pathway. These improvements should in turn improve the effectiveness, safety and quality of care for patients/consumers.

General Practice:

Pathology services contribute to all branches of medicine. They assist the clinical decision-making process and make a major contribution to the well-being of patients. This project will inform PHNs, which will in turn provide general practitioners (GPs) with variation data and benchmark data regarding the quality of pathology referrals in Australia. Such information will enable GPs to reflect on their own pathology referral practices and, where significant deviations from best practice care are identified, adjust their practices to improve the quality of care that they deliver to their patients.

Pathology laboratories:

Pathology laboratories provide a critical component of patient care. This project will contribute important findings that can enhance the test management process in general practice, so as to enhance the use of best practice and evidence-based guidelines.

Primary Health Networks:

The analysis of the quality of pathology referrals at different sites allows PHNs to perform a comparison of the effectiveness of pathology services across their respective networks. The benchmark data of pathology service utilisation at different sites can also be used by the PHNs to inform decisions for quality improvement by highlighting areas that may need additional support.

Government Departments of Health:

Departments of Health can use the benchmark results from the project to influence macro-level decision making and monitor the quality of pathology in general practice across different jurisdictions. This will contribute to major improvements in the quality of pathology and impact positively on patient outcomes.