Abstract
Abstract
Background
Evidence of significant variations in clinical biochemistry practice in Wales led to the formation of the All Wales Clinical Biochemistry Audit Group (AWCBAG) in 1993, with the aim of auditing laboratory services to ensure that they are optimally used. As part of this process, clinical guidelines are produced and circulated to all clinical biochemistry departments in Wales. The current aim of the AWCBAG is to assess the extent and impact of adoption of these guidelines across Wales.
Methods
Three surveys were despatched at intervals over a decade to all clinical biochemistry departments in Wales to investigate practice in: (1) urine albumin testing to screen for diabetic nephropathy; (2) biochemical investigation of menopausal status and the monitoring of hormone replacement therapy; (3) screening for Cushing's syndrome.
Results
The results show that laboratories across Wales are generally following guideline criteria and are adapting their practice in-line with changing recommendations.
Conclusion
The introduction of AWCBAG guidelines has been widely accepted by clinical biochemistry departments in Wales. These guidelines have led to a more efficient and effective use of laboratory services.
Introduction
Numerous investigations into clinical practice have demonstrated wide variations in the criteria used for the diagnosis of human diseases within the UK. 1 New interventions of proven value are not always readily adopted and practices of doubtful worth are not abandoned. Even when essentially the same diagnostic procedure is used, key details can differ significantly.
Discrepancies in practice have led to the development of clinical standards and guidelines, produced by a number of national and international organizations, such as the National Institute for Health and Clinical Excellence (NICE), the Scottish Intercollegiate Guideline Network and the World Health Organization. These guidelines are based, where possible, on the research-based evidence and provide recommendations to encourage uniform adoption of ‘best practice’ among health-care professionals, ensuring that high standards of care are achieved irrespective of where a patient is diagnosed or treated.
However, even in laboratory medicine, establishing whether guideline criteria are followed in routine practice is difficult. Some indication of the effectiveness of clinical guidelines in improving patient care and outcomes can be obtained by carefully designed audits that systematically review the performance against explicit criteria. 1–3
The All Wales Clinical Biochemistry Audit Group (AWCBAG) is an example of a regional audit project in clinical biochemistry, one of whose main aims is to narrow the evidence-practice gap by producing clinical guidelines. Since 1993, the AWCBAG has carried out many audits in Wales and produced 25 guidelines (Box 1).
Current Guidelines issued by the All Wales Clinical Biochemistry Audit Group
Analysis of cerebrospinal fluid for xanthochromia
Biochemical investigation of inherited metabolic disorders in children
Biochemical investigation of menopausal status and monitoring of hormone replacement therapy
Biochemical markers of myocardial damage
Investigation of macroprolactinaemia
Investigation of porphyria
Investigation of serum and urine paraproteins
Investigation of short stature in children
Laboratory investigation of polycystic ovary syndrome
Laboratory investigation of renal stone disease
Measurement of ammonia in blood
Monitoring glycaemic control in patients with diabetes mellitus
Multiple blood gas analysers in the same Trust
Performance and interpretation of the short synacthen (tetracosactide) test in investigating suspected adrenocortical insufficiency
Performance of the oral glucose tolerance test
Point-of-care testing
Provision and reporting of lipid analyses
Prostate-specific antigen measurement
Standards for screening for Cushing's syndrome
Sweat testing
The laboratory investigation of suspected phaeochromocytoma in adults
Thyroid function testing strategies
Trace element analyses and monitoring of total parenteral nutrition
Urine albumin testing in screening for diabetic nephropathy
Use of automated immunoassay analysers
Guidelines can be accessed at
The process of guideline production is as follows:
A relevant topic is chosen and a questionnaire to survey current practice is prepared by the project initiator and reviewed by the AWCBAG committee; The questionnaire is distributed to all clinical biochemistry laboratories in Wales; The project lead presents a summary of completed responses at an all Wales audit meeting to clinical biochemists and other stakeholders. Follow-up of non-responders to questionnaires has generally ensured response rates of 85–100%; A draft guideline recommending best practice is prepared by the project lead, reviewed by the AWCBAG committee and circulated for consultation to clinical biochemistry staff within Wales; Once agreed by laboratory staff, the guideline is then sent to the appropriate medical group (e.g. Welsh Endocrine and Diabetes Society) for their comments; The guideline is then finalized by the AWCBAG committee, distributed to laboratories across Wales for implementation and published on the AWCBAG website (
The current challenge for the AWCBAG is to assess the clinical effectiveness of this initiative. Selected guidelines have therefore been re-audited at intervals of 3–5 years as a way of undertaking continuous peer review of the practice of clinical biochemistry in Wales; practice has been compared with the standards in the published guidelines. The main aim of this study was to investigate whether the introduction of guidelines through the audit scheme has changed the current practice by evaluating the results from surveys of three topics for which there are current AWCBAG guidelines.
Methods
Questionnaires were distributed to all clinical biochemistry departments in Wales to investigate the current practice in three subjects for which standards have been issued by the AWCBAG. These were:
Urine albumin testing to screen for diabetic nephropathy; Biochemical investigation of menopausal status and the monitoring of hormone replacement therapy (HRT); Screening for Cushing's syndrome.
Current guidelines for these topics can be found on the AWCBAG website (
Three surveys were carried out for each topic, including an initial survey that assessed practice and methodology across Wales before the standards were developed. Two further post-standard audits were performed using questionnaires with a similar content to the original pre-standard one (Table 1). The questionnaires consisted of a number of multiple choice and one-word/sentence answer questions. Results obtained from the three audits for each topic were compared in order to assess the extent of changes in clinical practice during the period since introduction of the AWCBAG guidelines. Where given, statistical comparisons have been undertaken using Fisher's exact test.
Guidelines produced by the All Wales Clinical Biochemistry Audit Group investigated in this paper: year of initial survey, guideline production and subsequent re-audits
Results
Replies were received from nearly all Welsh clinical biochemistry laboratories for every survey despatched; the response rate exceeded 85% on each occasion and at least 12 replies were received for each questionnaire. The results from the three completed surveys for each topic (Table 1) provide a summary of clinical practice throughout Wales over a decade. The initial surveys identified a wide variation of practice between departments, with differences in choice of screening and monitoring protocols, analytical methods and reporting procedures for the same tests. Following the production of guidelines, subsequent audits showed more harmonization of practice. Key findings for each topic are reported in the following sections.
Standard 1: Urine albumin testing in screening for diabetic nephropathy
This guideline, originally written in 1998, was revised in 2002 to incorporate new guidance published by NICE for management of type 2 diabetes; 4 including gender-specific reference ranges for the albumin/creatinine ratio. However, although a timed overnight (or 24 h) urine collection is the gold standard test for assessing urine albumin excretion, 5 the Welsh guideline recommends the more practical measurement of the albumin/creatinine ratio in an early morning sample. The clinical utility of timed urine collections is limited by inconvenience to patients, inaccurate collections and the burden on laboratory staff processing specimens. Furthermore, there is a good correlation between the albumin (or protein) to creatinine ratio in early morning spot urine samples and timed 24 h or overnight urine collections. 6
In 1997, Welsh laboratories were using several different sample types to screen for microalbuminuria (Table 2), with four of the 12 laboratories using an early morning urine sample for the initial screening test, compared with 10 of 13 laboratories in the 2006 audit (P < 0.05). Interpretative ranges have also become more uniform: in 2001, only four of the 12 laboratories (4 of 8 who were using an early morning or random urine sample as the initial screening test) quoted the recommended cut-off value of 2.5 mg/mmol creatinine; whereas in 2006, 10 of the 13 laboratories (10 of 12 who were using an early morning or random urine sample) quoted the revised recommended cut-off values of 2.5 and 3.5 mg/mmol creatinine for men and women, respectively (P < 0.05).
Sample collection protocols for screening for albuminuria in Wales: number of departments that used each test at time of audit
*P < 0.05 compared with data for 1997 relative to the use of other screening tests
Standard 2: Biochemical investigation of menopausal status and monitoring of hormone replacement therapy
This guideline was finalized in 1998 and then circulated to clinical biochemists and gynaecologists in Wales. A key recommendation is that only follicle-stimulating hormone (FSH) is measured when initially investigating menopausal status in women aged 45 years or over; other recommendations concern tests for monitoring HRT.
The original survey in 1997 showed that most departments in Wales were measuring a number of reproductive hormones, including luteinizing hormone (LH) and oestradiol, in addition to FSH, to investigate menopausal status in women aged 45 years or older (Table 3). After the guideline was issued, more laboratories were using FSH alone for this purpose: seven of 13 in 2002 and six of 13 in 2006, compared with three of 15 in 1997, but these changes are not statistically significant.
Hormone profile recommended by clinical biochemistry departments in Wales to investigate the menopausal status of women aged 45 years or over: number of departments that used each test at time of audit
FSH, follicle-stimulating hormone; LH, luteinizing hormone
More generally, the initial survey in 1997 showed that Welsh laboratories (n = 15) were using 54 and 31 different sets of hormone tests to investigate menopausal status (in various age groups) and the monitoring of HRT, respectively, compared with 45 and 22 hormone sets, respectively (by 13 laboratories) in 2002, and 37 and 24, respectively (by 13 laboratories) in 2006; however, these changes are not statistically significant.
Standard 3: Screening for Cushing's syndrome
An initial survey of urine cortisol assays in 1995 led to a further survey in 1996 of the tests used to investigate suspected Cushing's syndrome. Guidelines were issued in 1998, which recommended the overnight dexamethasone suppression test, with a 1 mg dose as the first-line screening test. The 24 h urine cortisol measurement was also suggested as an alternative in adults, provided renal function was normal.
The 1996 survey showed that most laboratories offered random serum cortisol and diurnal cortisol variation (with samples for cortisol collected at 9:00 h and 24:00 h), as well as 24 h urine cortisol and the overnight dexamethasone suppression test, as first-line tests for investigating suspected Cushing's syndrome (Table 4). The later audits in 2001 and 2006 showed that laboratories are continuing to recommend the dexamethasone suppression test and 24 h urine cortisol for diagnosing this condition. In contrast, in 2001 the use of non-recommended random and timed serum cortisol measurements had declined significantly (P < 0.05) relative to the use of recommended screening tests, compared with the initial findings in 1996 (Table 4); however, in 2006 this decline was not quite significant (0.05 < P < 0.10).
Diagnostic tests recommended as first-line screening for Cushing's syndrome in Wales: number of departments that used each test at the time of audit (most used more than one)
*P < 0.05 compared with data for 1997 relative to the use of recommended screening tests
In 1996, the dose used in the overnight dexamethasone suppression test varied, with three of 12 Welsh laboratories using a 2 mg rather than a 1 mg dose of the steroid. The guideline recommends a 1 mg dose of dexamethasone. By 2001, 12 out of 13 laboratories were using a 1 mg dose and the most recent audit in 2006 showed that there is now 100% compliance with this criterion.
Discussion
The results from the audits performed in 2001–2002 and 2006 demonstrate increasing adherence to the recommended guidance with consequent greater harmonization of practice across Wales. However, not all the changes are statistically significant, probably due to the relatively small numbers of laboratories (12–15) surveyed.
For diabetic nephropathy, by 2006 nine of the 12 surveyed laboratories were measuring the albumin/creatinine ratio in an early morning urine sample to screen for and monitor this condition and using the recommended cut-off values. The use of timed overnight urine collections has gradually decreased from 1997 to 2006.
For biochemical investigation of menopausal status and HRT monitoring, the number of test combinations used has decreased, but disappointingly in both 2002 and 2006, only about half of laboratories were adhering to the recommendation that FSH only is sufficient for initial investigation of menopausal status in women aged 45 years or over. Possible explanations include clinicians' reluctance to change practice, use of fixed test combinations to simplify laboratory operating procedures and insufficient qualified staff to vet requests. Use of ‘tick box’ compared with ‘free text’ request forms has not been formally surveyed within Wales, but several Welsh laboratories still use ‘tick box’ request forms with an FSH/LH combination option.
The gradual shift towards exclusive use of the overnight dexamethasone suppression test, employing a 1 mg dose, means that there is now more consistency in testing for Cushing's syndrome throughout Wales. Uniformity in practice is particularly important for this condition, given that it is a relatively rare disorder and patients are generally referred to specialist endocrinologists in regional centres for further management. However, a few laboratories still used non-recommended random and timed serum cortisol measurements to screen for Cushing's syndrome in both 2001 and 2006.
There has been a significant change in practice since the introduction of the guidelines for some recommendations. However, this was not true for all the recommendations that were re-audited, and there was less change between the audits in 2001 and 2006 in comparison with the initial changes following introduction of the guidelines. It is also uncertain whether the guidelines themselves are responsible for the changes in practice, or whether there have been other influences, e.g. from guidelines produced by NICE and/or other professional groups. Even where the changes were significant, not all laboratories have altered their practice, possibly due to clinicians' reluctance to do so, and changes in practice cannot be enforced unless guidelines are mandatory. However, the opinions of the groups of clinicians initially consulted on the guidelines will be sought about the results of the re-audits.
Conclusion
The main aim of this study was to determine whether the introduction of clinical biochemistry guidelines has changed current practice in Wales, by performing an initial survey before and subsequent audits after the introduction of various clinical biochemistry guidelines. This process has shown that, in part, practice has improved and become more uniform since publication of such guidelines.
Implementation of best practice and increased harmonization are both important objectives to ensure that all patients receive the best possible care wherever they live and to simplify the task of clinical staff caring for patients from different localities, e.g. in managed networks. This study has shown that clinical auditing and the resulting production of standards/guidelines is a worthwhile method of improving the quality of service provided by the clinical biochemistry departments in Wales. It has helped reduce the number of unnecessary tests although the methodology did not allow us to quantitate this and allow consideration of cost savings. New ways of working are constantly being introduced and this process also provides an effective way of keeping practice up to date on a regional basis.
Footnotes
Acknowledgements
We thank the following for their contributions to the guidelines and audits discussed in this paper:
Current and former committee members of the AWCBAG; Topic leads: Mrs M A Thomas (urine albumin testing), Mr J A Tovey (HRT and menopause investigations) and Dr G F Read (screening for Cushing's syndrome); The wider clinical biochemistry community in Wales for completing questionnaires; Members of the Welsh Endocrine and Diabetes Society for their contributions to the guidelines on urine albumin testing and screening for Cushing's syndrome.
This article was prepared under the auspices of the AWCBAG. Dr K D Griffiths has been chairman since its inception in 1993 and Dr D A Oleesky was secretary from 1995 to 2005.