Clinical audit in the laboratory

Transcription

1 Department of Chemial Pathology, National Health Laboratory Servie, Tygerberg Hospital, University of Stellenbosh, Cape Town, South Afria Correspondene to: Professor R T Erasmus, Department of Chemial Pathology, National Health Laboratory Servie, Tygerberg Hospital, University of Stellenbosh, PO Box 19113, Tygerberg 7505, Parow, South Afria; rte@sun.a.za Aepted 10 February 2009 Clinial audit in the laboratory R T Erasmus, A E Zemlin ABSTRACT Audits are part of the ontinuous quality improvement proess and one of the key elements of linial governane. Laboratory-based linial audits are onerned primarily with the everyday aspets of laboratory servies and are a means of providing feedbak to the users of the laboratory and its staff. They involve measuring the performane of laboratory servies against established standards. These standards have ideally been established using the priniples of evidene-based mediine. If neessary, hanges are implemented and then a re-audit is performed after a ertain time period to ensure that the hanges have been implemented and maintained. Areas of audit in the laboratory inlude the preanalytial, analytial and postanalytial phases. This review artile examines the basis of linial audits in the laboratory and then proeeds to desribe in detail how a laboratory-based linial audit should be performed and monitored, with speial referene to the hemial pathology laboratory. Healthare delivery organisations globally are utilising various quality indiators to measure the effiay of speifi interventions as well as to identify healthare improvement opportunities. These quality indiators are further being used for performane and outome measurements as a means to measure, monitor and improve the quality of are and servies. Laboratory data are important in the medial deision making proess and influene 70% of medial diagnoses. 1 Audit means to evaluate, and in the ontext of the pathology laboratory would mean a systemati and ritial analysis of pathology servies. An audit is a quality improvement proess and is an essential part of the quality assurane programme of a laboratory. The standard definition of linial audit is a quality improvement proess that seeks to improve the patient are and outomes through systemati review of are against expliit riteria and the implementation of hange. 2 Five distint ativities an be onsidered under the broad umbrella of audit: Solving problems assoiated with proess or outome Monitoring workload in the ontext of ontrolling demand Monitoring introdution of new tests and/or hanges in pratie Monitoring adherene with best pratie (eg, with guidelines) 3 Monitoring of analytial quality. HISTORY Audit is not a new proess. 4 As early as 1750 BC, King Hammurabi, the 6th king of Babylon, instigated audit for liniians. 2 In modern mediine, one Review of the first linial audits was undertaken by Florene Nightingale during the Crimean War of ; she applied strit sanitary routines and hygiene standards that dereased the mortality rates from 40% to 2%. Another famous figure who advoated linial audit was Ernest Codman ( ), an orthopaedi surgeon at Harvard Medial Shool. He beame known as the first true medial auditor following his work in 1912 on monitoring surgial outomes. Despite the early work of these pioneers, linial audit is relatively new to modern medial praties. CLINICAL GOVERNANCE AND AUDITS Clinial audit is one of the key elements of linial governane, the latter being a system through whih healthare organisations are aountable for ontinuously improving the quality of servies. It is desribed as a framework through whih organisations are aountable to ontinue to improve the quality of the servie and safeguard high standards of are by reating an environment in whih exellene in linial are would flourish. 5 In 1989, the White Paper Working for Patients saw the first move in the UK to standardise linial audit as part of professional healthare, 6 and later it was stipulated that time was to be alloated for audit work within eah onsultant s job. 7 In an attempt to assess linial performane, medial audit was introdued as part of the 1997 white paper health reforms of the NHS in the UK. 8 In 1997, the Royal College of Pathologists published guidelines entitled Clinial Audit in Pathology 9 and established The Professional Standards Unit for providing guidane to pathologists to produe evidene on the quality of servie they provide. 10 In 2005, the Royal College of Pathologists published a ode of pratie for linial biohemists and linial biohemistry servies in whih linial biohemists were required to partiipate in linial audits to assess the quality and appropriateness of the servies provided. 11 Benhmarking shemes, whih share similarities with linial audits, have been in existene in the USA for many years. Benhmarking is the proess of measuring produts, servies and praties against leaders in a field, allowing the identifiation of best praties that lead to sustained and improved performane. AUDIT AND EVIDENCED-BASED MEDICINE Evidene-based mediine (EBM) is at the ore of ontinuous quality improvement programmes and audits are very muh part of the ontinuous proess to implement and maintain best pratie in the laboratory. Audits use EBM to set the standard. All of the audit ativities are found in the pratie of evidene-based laboratory mediine J Clin Pathol 2009;62: doi: /jp

2 (EBLM), namely that there is a linial question for whih the test result should provide an answer and that the answer will lead to a deision being made and an ation taken, leading to an improved health outome. EBM and EBLM are essential tools in the assessment of linial effetiveness, as high quality systemati linial researh is neessary for investigating the impat of any intervention on linial outome. 12 Fig 1 shows how EBLM and audit are linked. 13 AUDIT VERSUS RESEARCH 2 Aording to the Researh Governane Framework for Health and Soial Care, researh an be defined as the attempt to derive new knowledge by addressing learly defined questions with systemati and rigorous methods. 14 In general, researh generates evidene to support, refute or develop a hypothesis. It aims to study the effet of hange on linial or servie pratie. Clinial audit, on the other hand, aims to improve patient outomes by improving professional pratie and the general quality of servies rendered. It ompares urrent pratie against a standard that has already been set and examines whether this pratie meets required standards, follows published guidelines and applies the knowledge that has been gained through researh. In some instanes there might be no standard praties that have to be developed. An important distinguishing differene between the two is that standard of praties are the basis of measurement in linial audits and not hypotheses (table 1). Researh may diret audit, whih may lead to the development of hypotheses that an be tested by researh. 2 Researh inreases overall knowledge and disovers best pratie. Researh an identify areas for audit that are ritially important for patient are. As resoures are limited in the laboratory, researh an identify areas that need to be prioritised and monitored. Audits may our as a final step for a good laboratory-based researh programme, and they an pinpoint areas where the researh evidene is laking. In a systemati review of studies that have examined inappropriate laboratory use, it was found that many of these studies did not use impliit or expliit riteria, thus not meeting methodologial standards, and suggesting that researh is needed to develop evidene-based standards for measuring inappropriateness of laboratory test use. This may be a hallenging task, as there an be differing views on the appropriateness of a test request. When inappropriate laboratory Figure 1 Evidene-based laboratory mediine and audit. 17 utilisation is identified, methods to retify the problem must be developed, tested and implemented. STEP-BY-STEP APPROACH TO PERFORMING AN AUDIT OF LABORATORY SERVICES Ideally, linial audits should be onduted within a strutured programme, with effetive leadership, involving all staff, and with emphasis on team work and support. Appropriate funding should be made available, as linial audits may involve substantial osts, 80% of whih are related to the use of staff resoures. 15 Time is alloated for personnel to partiipate in linial audits, as they an be time onsuming, and funding is disussed. Given the signifiant levels of resoure input, researh is needed into the value of linial audit, and the ost impliations of linial governane need to be expliitly reognised. Some institutions reommend that all audits should be registered and approved by the institutional ommittee. Often, audit results in ritiism of other departments or individuals without their knowledge or involvement, thus joint audit is desirable and should be enouraged. The audit yle Clinial audit is a ylial proess (fig 2). 16 Coles suggested ombining the learning yle and the audit yle to produe a double-looped yle starting with observing pratie, refletion on pratie, setting new standards and, after further observation and refletion, implementing hange followed by monitoring and re-audit. 17 The stages of a linial audit There are five stages of a linial audit. 16 Stage 1 This involves preparation for the audit, inluding hoosing a team, o-ordinator and a topi. Ideally, the topi should be relevant with potential benefit to patient are or the organisation. Three areas an be addressed: preanalytial, analytial and postanalytial (box 1). A quantity impat analysis may be performed to hoose and prioritise ideas for a linial audit. 19 This produes a list of topis whih an be prioritised. Another approah is to examine the number of points on test request pathways for speifi linial onditions. At eah point there are aspets of the testing proess that an be audited. Stage 2 This involves defining and seleting the riteria and standards to whih the performane will be ompared. A riterion is a speifi statement of what should be happening, is used to assess quality of are and is preferably evidene based; therefore a literature searh is essential. 20 Criteria are often also alled guidelines or benhmarks and are based on biomedial researh and health tehnology assessment. They are outome orientated and usually developed by a multidisiplinary team. 21 Unfortunately, the quality of published guidelines is variable, many not being expliit or evidene based. 18 If no aeptable guidelines are available, the organisation needs to develop their own best pratie guidelines before the linial audit is ommened. The All Wales Clinial Biohemistry Audit Group has published 25 guidelines for various investigations in linial biohemistry. 22 Their group reently evaluated the impat of these guidelines on laboratory servies in their area of pratie 594 J Clin Pathol 2009;62: doi: /jp

3 Box 1 Areas of audit in the laboratory 18 Preanalytial Request forms; are they easy to use? Are all relevant details provided by the user? Speimens: is the right speimen reeived at the right time? Are the appropriate investigations seleted by the laboratory staff Phlebotomy servies and transport of samples to laboratory Analytial Is the range of investigations available appropriate? The number of requests for a speifi test and the positivity rate should be audited. Those tests for whih requests whih are rare and/or have a low positivity rate should be withdrawn Are the test methods being arried out aording to standard operating proedures? Safety poliies and proedures. Every laboratory should have a omprehensive safety poliy. Every single aident in the laboratory should be reorded and improvements made if neessary. The use of dangerous substanes should be audited Effiient use of staff. The training of all staff may be audited. Purhasing of equipment, reagents, stationary and other items Laboratory reports: are they preise and lear? Storage of reagents and speimens Internal and external quality assessment Test utilisation Postanalytial Turn-around times for eah request. Attempts should be made to monitor the turn-around time in eah department and see whether improvements an be made Reporting methods (eg, types of reports, diret ommuniation, omputer system) Referene ranges Interpretation, onsultation and omments on reports Complaints and orretive ation taken (Wales) and found that they had led to a more effiient and effetive use of laboratory servies. 23 Stage 3 This involves data olletion and measuring the performane. The data should be preise and relevant, and olleted over a speified time period. Patient registers may be used to identify patients, and linial reords are often used as data soures. Audit staff needs to be areful about the auray, timeliness and ompleteness of linial reords. The data olleted must be adequate (sample size) and relevant to enable making valid onlusions. 24 If a questionnaire is used, it should be lear, onise and unambiguous, avoiding double-barrelled and leading questions. Data are generally olleted retrospetively and most audits only require the use of basi desriptive statistis. 25 Where possible, the results should be ompared with audits performed by peer groups or national standards. 26 Stage 4 This involves analysing the results obtained and making improvements, identifying barriers and obstales to hange. An integrated plan is developed for the delivery and monitoring of interventions. Implementation methods that are most Table 1 Audit Some differenes between audit and researh Researh Determines whether the laboratory is following agreed pratie Measures against aepted standards Reviews urrent pratie and ompares it with best pratie Creates new knowledge and provides foundations for national and/or loal agreement about what kind of linial are and treatment we should be providing Based on a hypothesis Improves overall knowledge and disovers best pratie suitable are disussed and may involve eduation of laboratory staff and users of pathology servies using outreah programmes, presentations and group disussions. Stage 5 This involves a re-audit after an agreed time period to ensure that improvement is sustained. A single data olletion does not onstitute an audit, 36 as the first data olletion establishes the laboratory s urrent position and the seond one establishes if improvements have been made. 26 A timesale for a reaudit is deided upon using the original design. 27 Any hanges that have improved servie delivery are monitored, evaluated, sustained and reinfored in a supportive environment. This stage is ritial to the suessful outome of the audit proess, as it determines if hanges implemented have resulted in an improvement of the laboratory servies. Finally, the audit report is written in a strutured way with a title and bakground to the study. Aims and objetives should be learly stated, defining the riteria and standards used and ending the report with a onlusion and reommendation for improvement and reaudit. 27 Results of a good audit should be disseminated loally via the loal health authorities and nationally through professional organisations and meetings. The Royal College of Pathologists has published examples of high quality laboratory-based linial audits under the title Writing your audit report. 27 IMPACT OF AUDIT ON LABORATORY SERVICES As pathologists, our obligation to linial mediine is to monitor test usage and to evaluate their usefulness. Clinial audits in the laboratory are onerned primarily with the daily aspets of laboratory servies and are useful for providing feedbak to the users of the laboratory and its staff. They have been benefiial in assessing and modifying laboratory and linial pratie ranging from the use of servies to the linial effetiveness of laboratory tests as well as their impat on linial outomes. 28 Clinial are Clinial guidelines of a high quality may help liniians to hange the ordering pattern of tests to one that is more rational with a better ost/benefit ratio, the ultimate goal being to improve the quality of are of patients. 29 Van der Weiden et al studied the appropriateness of holesterol testing in their setting and found that holesterol testing was not being performed aording to national guideline reommendations. 30 Audits an be used to improve the quality of laboratory servie by dereasing turnaround time by improving various proesses involved in the testing proess, suh as sample registration time, test result validation and result delivery time. In a survey that examined inompletely filled in laboratory request forms and their impat on the phoning out of ritial results in our laboratory, we determined that they are detrimental to patient J Clin Pathol 2009;62: doi: /jp

4 Figure 2 The audit yle. 25 are and this is an example of how preanalytial errors may influene the linial ourse of a patient. 31 Quality assurane programmes Clinial audits an highlight areas where better use of resoures an be implemented. 32 They may also be used to assess whether quality assurane methods in the laboratory are adequate and onform to aepted standards. Housley et al reently desribed an audit of the use of internal quality ontrol proedures in 54 laboratories and highlighted signifiant variability in internal quality ontrol pratie. They onsequently proposed a set of regional standards that were developed and disseminated and a reaudit at a future date has been proposed. 33 Audits an be used to provide evidene of poorly managed external quality assessment shemes or programmes that are too analytially foussed. Test utilisation In an effort to ut osts, healthare organisations are studying test utilisation and its impat on linial outomes. Clinial audit has been used to examine the requesting patterns and identify problem areas. Mainwaring et al performed an audit of oagulation sreen requests of patients admitted to the medial assessment unit and found that numerous tests were performed unneessarily leading to an unneessarily high workload and unneessary osts. 34 An audit by the North Thames Clinial Chemistry Audit and QA Group in 2004 highlighted several problem areas in test requests for aldosterone and renin. These ranged from improper proessing and storage to inadequate preparation of patients and inappropriate use of referene ranges. 35 In our laboratory we observed that although phosphate is an important indiator of mortality in espeially the seriously ill patient, it is not a regularly performed test. 36 Similarly, Kwok and Jones arried out an audit on unneessary repeat testing in an immunology laboratory and found that this ontributed to high ost and a waste in tehniian time and reagents. During a 12 month period, they found that repeat requests for tumour markers and autoantibodies made up 16.8% of the workload leading to waste of resoures. 37 Audits have been reported to improve ommuniation among olleagues leading to improved patient are, inreased professional satisfation and better administration, as illustrated by an extensive review of 93 publiations by Johnston et al. 38 Ultimately it is the pathologist s duty to help liniians order the appropriate tests, at the orret time, in the orret order. Combination of pratie guidelines, modifiation to the laboratory requisition form and funding poliy hanges have been found to signifiantly derease the use of several tests. 39 MONITORING FOR A SUCCESSFUL CLINICAL AUDIT Ideally, there should be an audit ommittee in the organisation in harge of all linial audits. This ommittee should have a lear strategy with speified programmes and assoiated ativities. They form a support group, organise funds and meet regularly to monitor the progress of. An up to date database of audits in progress should be readily available. All staff should partake and reeive adequate training. The onlusions and riteria of the audits should be standard based aording to EBM priniples. There must be evidene of appropriate ation plans that have been implemented aording to the results found. CONCLUSION Systemati improvement of health servies requires the objetive measurement of people, praties and organisations against valid and expliit standards in order to identify and implement appropriate hange. 40 Laboratory-based linial audits are an essential part of total quality management. They are important for good linial governane. In the laboratory they have been found useful in developing guidelines for testing, assessing test utilisation and various aspets of quality assurane programmes. 596 J Clin Pathol 2009;62: doi: /jp

5 Interative multiple hoie questions This JCP review artile has an aompanying set of multiple hoie questions (MCQs). To aess the questions, lik on BMJ Learning: take this module on BMJ Learning from the ontent box at the top right and bottom left of the online artile. For more information please go to: jp.bmj.om/eduation Please note: the MCQs are hosted on BMJ Learning the best available learning website for medial professionals from the BMJ Group. If prompted, subsribers must sign into JCP with their journal s username and password. All users must also omplete a one-time registration on BMJ Learning and subsequently log in (with a BMJ Learning username and password) on every visit. Take-home messages Audits are part of ontinuous quality improvement in the laboratory and one of the key elements of linial governane. Audits ompare urrent pratie against an aepted standard that has been set and examine whether this pratie meets the standard. Audit is a ylial proess: it ompares pratie to standards, measures performane, makes improvements and, importantly, involves a re-audit after a time period to ensure that the improvement is sustained. There are ample areas for audit in the laboratory, ranging from the preanalytial to the postanalytial phases. Additionally, they forge important liaisons between linial and laboratory staff, leading to improved ommuniation. They ultimately lead to improved patient are and ost ontainment. Competing interests: None. REFERENCES 1. Quality Institute. Making the laboratory a key partner in patient safety. Proeedings of the Quality Institute Conferene; April 2003, Atlanta (GA). 2. Copeland G. A pratial handbook for linial audit. London: NHS Clinial Governane Support Team, Burtis CA, Ashwood ER, Bruns DE, eds. Introdution to laboratory mediine and evidene-based laboratory mediine. In: Tietz textbook of linial hemistry and moleular diagnostis. 4th edn. St Louis (MO): Saunders, 2006: Anon. Clinial audit. (aessed 16 Marh 5. Trenti T, Canali C, Sognamiglio A. Clinial governane and evidene-based laboratory mediine. Clin Chem Lab Med 2006;44: Department of Health. Working for patients. London: Stationary Offie, Department of Health. Terms and onditions of servie of hospital and medial and dental staff. London: Stationary Offie, Waise A. Clinial audit and the ontribution of the laboratory to linial outome. Clin Chim Ata 1999;280: The Royal College of Pathologists. Clinial audit in pathology. London: The Royal College of Pathologists, The Royal College of Pathologists. Clinial governane and revalidation: the role of linial audit. Edu Pathol 2002;117: Beastall G. Code of pratie for linial biohemists and the linial biohemistry servies. London: The Royal College of Pathologists, Horvath A R, Watine J, Miko TL. ISO 15189:2003 and evidene based laboratory mediine. ejifcc 2004;15(4). (aessed 6 April 13. Prie CP. Evidene-based laboratory mediine: supporting deision making. Clin Chem 2000;46: Department of Health. Researh governane framework for health and soial are. 2nd edn (aessed 16 Marh, 15. Lok P, MElroy B, Makenzie M. The hidden ost of linial audit: a questionnaire study of NHS staff. Health Poliy 2001;51: National Institute for Clinial Exellene. Priniples for best pratie in linial audit. London: National Institute for Clinial Exellene, Coles C. Making audit truly eduational. Postgrad Med J 1990;66(Suppl):S Plebani M. The importane of laboratory reasoning. Clin Chim Ata 1999;280: NHS. University Hospitals Bristol. (aessed 16 Marh 20. Verma R. Clinial audit. Anaesth Intensive Care 2004;5: Oosterhuis W, Bruns D, Watine J, et al. Evidene-based guidelines and priniples in laboratory mediine: priniples and methods. Clin Chem 2004;50: All Wales Clinial Biohemistry Audit Group. (aessed 16 Marh 23. Griffiths KD, Duffy EJ, Oleesky DA. Clinial biohemistry guidelines in Wales: their impat on laboratory servies. Ann Clin Biohem 2008;45: Siaovelli L, Sehiero S, Zardo L, et al. External quality assessment: an effetive tool for linial governane in laboratory mediine. Clin Chem Lab Med 2006;44: UBHT Clinial Audit Central Offie. How to analyse audit data. uhbristol.nhs.uk/douments/how_to_analyse_audit_data.pdf (aessed 16 Marh 26. Verma R. Data quality and linial audit. Anaesth Intensive Care 2006;7: The Royal College of Pathologists. Writing your audit report. org/index.asp?pageid=1451 (aessed 16 Marh 28. Jakson BR. Managing laboratory test use: priniples and tools. Clin Lab Med 2007;27: Drife JO. Perinatal audit in low- and high-inome ountries. Semin Fetal Neonatal Med 2006;11: Van der Weijden T, Danson A, Shouten B, et al. Comparison of appropriateness of holesterol testing in general pratie with the reommendations of national guidelines: an audit of patient reord in 20 general praties. Qual Health Care 1996;5: Nutt L, Zemlin AE, Erasmus RT. Inomplete laboratory request forms: the extent and impat on ritial results at a tertiary hospital in South Afria. Ann Clin Biohem 2008;45: Cutinha PE, Potts KL, Rosario DS, et al. A prospetive audit of the use of a prostate lini. Br J Urol 1996;78: Housley D, Kearney E, English E, et al. Audit of internal quality ontrol pratie and proesses in south-east of England and suggested regional standards. Ann Clin Biohem 2008;45: Mainwaring CJ, Chana J, Evans J. Best pratie in audit: audit of oagulation sreen requests from patients admitted to the medial assessment unit. London: The Royal College of Pathologists, Kyd P. Audit of aldosterone and renin investigations. London: North Thames Clinial Chemistry Audit and QA Group, Hoffman M, Zemlin AE, Meyer WP, et al. Hypophosphataemia at a large aademi hospital in South Afria. J Clin Pathol 2008;61: Kwok J, Jones B. Unneessary repeat requesting of tests: an audit in a government hospital immunology laboratory. J Clin Pathol 2005;58: Johnston G, Crombie IK, Davies HT, et al. Reviewing audit: barriers and failitating fators for effetive linial audit. Qual Health Care 2000;9: van Walraven C, Naylor CD. Do we know what test utilization is? A systemati review of laboratory lini audits. JAMA 1998;280: Shaw CD. Measuring against linial standards. Clin Chim Ata 2003;333: J Clin Pathol 2009;62: doi: /jp