Introduction

Recent media reports on the quality of hospital care inAustralia, and the establishment of Commonwealth andState committees on Quality in Health Care designed toreview medical performance (the Australian Council forSafety and Quality in Health Care; the National ExpertAdvisory Group on Safety and Quality in AustralianHealth Care, 1999; New South Wales Ministerial Advi-sory Committee on Quality in Health Care), indicate theneed for the Australasian Society of Cardiac and Thor-acic Surgeons (ASCTS) to urgently respond. The ASCTSmust define performance indicators, and compile localand national results of clinical performance in the prac-tice of cardiothoracic surgery.

Such data can then be used to measure the quality ofcare, and minimise inappropriate variations in practice,by allowing cardiothoracic surgical units and individualsurgeons to compare performance against best practicestandards at national and international levels. Import-antly, information can be acquired by trainees aboutmonitoring and practising quality improvement and onimproving communication. The Peer Review Committee(PRC) of the ASCTS is required to produce a frameworkfor monitoring and managing quality in cardiac surgicalpractice in Australia. The New South Wales HealthDepartment Steering Committee on Quality in Healthhas developed a six-point policy framework to ensuresafe, effective, appropriate, accessible and efficient healthtreatment and care with consumers as active partici-pants.1

Optimum quality management processes involveresponsibility shared between practitioners, units andinstitutions; the concept of shared clinical governance.

Peer Review and Quality Assurance in Cardiac Surgery in Australia

Alan W. Gale, FRACS

Peer Review Committee and National Cardiac Surgery Database Project

There is increasing community and government interest in determining the quality of medicalperformance in Australia. This means that any lead taken by the Australasian Society of Cardiacand Thoracic Surgeons (ASCTS) in defining and adopting performance indicators for theprofession is appropriate and timely. The ASCTS Peer Review Committee has examined theissue of quality management, and presents here several key indicators and a draft program ofremediation which, if adopted, will facilitate the achievement of the highest and most uniformsurgical standards for Australia. The Committee recommends the monitoring of just oneprocedure initially: coronary artery surgery. The Committee sees the quality managementprogram as involving both surgeons and hospital-based quality assurance teams. Ultimately, theprofession will gain from performance measurement by being able to minimise inappropriatevariations in practice. The community will benefit by having national and international bestpractice in cardiothoracic surgery, with surgeons being able to identify more easily thoseprocedural changes that result in superior as well as inferior outcomes. (Heart, Lung andCirculation 2001; 10 (Suppl.): S10–S13)

Key words: performance indicators, quality management.

Correspondence: Alan W. Gale, 1/341 Victoria Avenue, Chatswood,New South Wales 2067, Australia. Email: alangale@ozemail.com.au

Since the Bristol affair, hospital administrators arebecoming increasingly responsible for quality care andassessments of performance. In the UK, legislated clinicalindicators have been applied to cardiac surgical unitsunder a national performance framework that uses theabove six elements of health treatment. Regular auditsare now mandatory for all practising cardiothoracic sur-geons, with a warning that ‘fellow professionals couldthen provide extra training, supervision and support tocorrect what had been going wrong’ if outlying perform-ance is demonstrated.2 The ASCTS must avoid theseexternal constrictions by providing its own processes for‘policing its own’.

The peer review process is a component of qualitymanagement and must be ‘linked into the broader qual-ity improvement processes that involve the whole of and the highest levels of organisation in which care isdelivered’.3 Peer review committees must therefore workwith hospital quality assurance departments, dischargeplanners and clinical governance systems. The PRC ofthe ASCTS has drawn up a series of draft performanceguidelines for discussion by members of the Society.

Cross-dimensional issues are important componentsof performance assessment.4 The clinical competence ofindividuals or of a multidisciplinary care team, as well asthat of the hospital, is an issue that relates to safety, butdoes not fall directly under the duties of the PRC. Issuessuch as recertification and the conferring of credentialsare more properly the domain of the Royal AustralasianCollege of Surgeons (RACS), but may overlap with theinterests of the PRC and the ASCTS.

The Role of Peer Review

It is recognised that avoidable, unintended outcomesproducing harm are likely in the acute hospital environ-ment. ‘Error is an inevitable accompaniment of thehuman condition, even among conscientious profession-als with high standards. Errors must be accepted as evidence of system flaws and not character flaws.’5

Corrective efforts must avoid a culture of individualblame, but rather identify the system failures contribu-ting to the injury. It is accepted that unintended out-comes may not necessarily represent errors in judgementor technical performance.

Defining Key Performance Indicators

Key performance indicators (KPI), and indicators ofadverse events or of favourable results, provide the bestquality assessment tools and a basis for accountability.‘Key performance indicators are objective measures ofeither the process or outcome of patient care in

quantitative terms. They are not a direct measure ofquality and are simply intended to raise questions andprompt further investigation.’6 Key performance indica-tors are valid only if properly risk stratified (the functionof the ASCTS Data Review Committee), and if defini-tions are followed strictly.

In defining KPI, the PRC recommends concentratinginitially on one surgical procedure, coronary arterysurgery, which has a large denominator from whichmeaningful comparative statistics may be derived. ‘It ispreferable to collect data on a few relevant indicators asopposed to collecting a large number of irrelevant indica-tors simply to fulfil reporting requirements.’6

The optimal environment for discussing clinical resultsis the institutional peer review committee, and no sur-geon today can justify a failure to have his or her resultsregularly reviewed by colleagues. Such review currentlyconstitutes a component of recertification applications.

Sentinel Events

In the quality-care literature, sentinel events are high-lighted. ‘A sentinel event may be identified when levelsof performance patterns or trends vary significantly andundesirably from those expected or vary significantlyfrom that of other organisations or from recognised standards.’7 Sentinel events need investigation at thelevel of the institutional peer review committee at thetime of their occurrence, and do not necessarily fallwithin the province of investigation by the ASCTS PRC,unless they occur repeatedly.

Important Key Performance Indicators

Preoperative

The education of patients and relatives regarding majorrisks and postoperative expectations, as well as alter-native available treatment, forms the core of informedconsent. Nurses in most hospitals are already collectingpatient satisfaction data for hospitals and third partyinsurers, and unquestionably these data will becomeimportant in medico–legal issues. Communication withpatients and their relatives following unanticipated outcomes is even more important. While this is an indis-tinct concept for quantitative measurements, the ASCTSneeds to define a minimum standard in this area. Doctor–patient relationships are regularly assessed byhealth care complaints commissions, area health boards,medical registration boards and plaintiff lawyers. Arecent report in The Australian and New Zealand Journalof Surgery revealed patient priorities in informationgathering.8 The importance of surgeons providing

Heart, Lung and Circulation 2001; 10 (Suppl.) A.W. GalePeer review in cardiac surgery

S11

extremely difficult. Employing the traditionally accept-able range of two standard deviations each side of amean performance value, would identify, by definition,more than one or two surgeons outside these limits atany given time. Considerable effort has been made, particularly by the Victorian group, to defining moreappropriate margins of acceptable performance.

The requirements, by interested third parties, for regular indicative reports suggest that a minimum of12 months of data collection be made available. Anylesser time frame of analysis would be non-conclusive,but trends may be observed over a shorter time intervaland within departments, reviews every 6 months are recommended.

Practitioner Impairment

In the Australian context, it is unlikely that major impair-ments from substance abuse, psychiatric disease or crim-inal behaviour will be identified. Processes exist withinevery state jurisdiction for reporting and rehabilitatingsuch identified practitioners.

Suboptimal Performance

Suboptimal performance may occur in the absence ofearly diagnosed impairment, occurring as secondary tocharacter flaws, personal crises or failure to keep pacewith contemporary directions. Equally, over-applicationof surgical procedures, unsubstantiated by evidence,may constitute inadequate performance. Detection ofsuch issues and counselling must be an individualisedprocess for which no formal guidelines can be produced,and is best carried out within the individual unit.

It must be recognised that a national system for themandatory reporting of adverse events is imminent.9 TheNew South Wales Parliament is currently assessing aReport on Mandatory Reporting of Medical Negligencefrom the Committee on the Health Care ComplaintsCommission.10 It will soon be compulsory to report suspicions of negligent performance to the Health CareComplaints Commission. There is an urgent need for theASCTS to have in place its own performance-assessmentand remediation tools at a national level.

Although peer review committees may request Qual-ified Legal Privilege for quality improvement programs,in New South Wales this privilege has to be regularlyreviewed.11

Suggested Remediation Processes

Recurrent grade 1 deficiency requires intervention by the ASCTS. Investigations should be carried out with the

Heart, Lung and Circulation 2001; 10 (Suppl.)A.W. GalePeer review in cardiac surgery

S12

explanations to patients regarding alternative treatments,common risks and complications of the operative tech-nique in question can be seen in the high ranking givento such explanations in patient satisfaction surveys.

Equally, the decision on the most appropriate proce-dure, for example off-pump coronary artery bypass(OPCAB) versus open coronary artery surgery, should beused as an indicator and would probably be bestassessed as an intraoperative KPI.

Intraoperative

The frequency of use of the left internal mammaryartery compared to vein to the left anterior descendingcoronary artery as graft material can be used as a KPI.For now, duration of cardiopulmonary bypass, cross-clamp time or the conversion of OPCAB to open proced-ures, in the early experience of OPCAB, should beexcluded from KPI data collection, but may be appropri-ate for later surveys.

Early Postoperative

Key markers for early postoperative assessment are: (i)risk-stratified operative mortality (30 day or in-hospital);(ii) properly defined perioperative myocardial infarction;(iii) return to the operating room for bleeding orischaemia; (iv) deep sternal wound infections necessita-ting re-exploration; (v) permanent neurological deficit;and (vi) excess blood product use. These are all appropri-ate and easily compiled indicators.

Late Postoperative

Key markers for late postoperative assessment are cardiac-cause readmission, and re-intervention for graft failureor missed native coronary artery revascularisation (a90 day time frame may be appropriate with atrial fibrilla-tion, pleural effusion or gastrointestinal bleeds beingexcluded).

Grading Performance Indicators

Stratifying KPI as grade 1 or grade 2 may help defineoutlying performance and the remediation process.Grade 1 defects would relate to excess operative mortal-ity, perioperative infarction, cerebral complications orpostoperative use of intra-aortic balloon pumps or ven-tricular assist devices. Grade 2 defects would be exces-sive deep wound infections, use of blood products orcardiac-cause readmission.

Identifying Outlying Performers

Even within the chosen, easily monitored KPI, the defini-tion of outlying performance in the Australian context is

Heart, Lung and Circulation 2001; 10 (Suppl.) A.W. GalePeer review in cardiac surgery

S13

surgeon remaining anonymous for the reviewers. Aseries of steps is essential. First, the data need verifica-tion (possibly by the ASCTS Data Review Committee orits Chair). Then, it may be necessary to bring in an inves-tigation team, preferably from interstate, and possiblyincluding the Chairs of the PRC and Data Review Com-mittee as well as the President of the ASCTS and theChair of the Cardiothoracic Board.

Investigation would involve the team reviewing defi-nitions and descriptors, risk stratification, reportingmethods and the possibility of gaming. At this stage ofthe investigation, hospital records may need to beaccessed, requiring approval from hospital administra-tion. The investigation now enters a legal minefield. Thesurgeon under investigation must give the team author-ity to review the data and indemnify the ASCTS for useof the information.

If outlying performance is not proven, the problemmay have been failure of adherence to strict parametersby data managers, or deliberate gaming. Once a problemis identified, a repeat review in 12 months would be allthat is necessary.

If underperformance is identified and proven, a seriesof steps could be recommended: (i) notify the surgeon;(ii) notify the head of the unit; (iii) notify the administra-tion of the hospital from which the data is derived, andalso possibly other institutions employing the particularsurgeon; (iv) counsel the surgeon (carried out by a com-mittee with representatives of both the ASCTS andRACS); (v) review results in 3 months; and finally (vi) ifthere is no improvement, notify the state medical boardand state health department.

It is considered extraordinarily unlikely that theassessment of any surgeon or surgical unit in Australiawould need to progress beyond steps ii and iii above. Ifsuperior outlying performance is identified, a datareview is required to ensure that over-reporting of highrisk cases or under-reporting of unfavourable outcomeshas not occurred. If true superior performance isproven, notification of superior techniques or protocolsshould be provided to all ASCTS members.

Conclusion

The necessity for leadership by the ASCTS in the data collection, analysis and evaluation of cardiac surgical out-comes in Australian surgical units has become urgent asexternal quality measures are being imposed at FederalGovernment and State Government levels. This paperaddresses the direction in which quality outcome is moving, and proposes a simplified format for providingASCTS members with a means of defining best practiceresults and initiatives in Australia. I hope that surgeonswill use this paper as a basis for construction of the opti-mal processes necessary to define best practice in our spe-cialty.

References

1. NSW Health Department. A Framework for Managing the Qualityof Health Service in NSW. Sydney: NSW Health Department;1999 (publication no. HPA 990100).

2. Dodson, Frank (UK Secretary of State for Health). Speech given9 June 1998. Available from URL: http://ccad3.biomed.gla.ac.uk/ccad/frank.html

3. Quality in Australian Health Care Taskforce. Final Report.Canberra; 1996.

4. Quality Framework for NSW Health Steering Committee.Quality and Budget: Equal Partners in Health. Sydney: NSWHealth Department; 1999 (publication no. HPA 990100).

5. Leape L. Error in medicine. JAMA 1994; 272: 1851–7.6. Northern Sydney Health Quality Kit (Royal North Shore

Hospital in-house document). June, 2000.7. Joint Commission on Accreditation of Healthcare Organizations.

Sentinel Event Alert Bulletin (12); 1999. Available from URL:http://wwwb.jcaho.org/edu_pub/sealert12.html

8. Courtney MJ. Information about surgery: What does the publicwant to know? Aust. N.Z. J. Surgery 2001; 71: 24–6.

9. Barraclough B. Quality in the face of adversity. Aust. Med. 2000;12: 14.

10. Committee on the Health Care Complaints Commission,Parliament of NSW. Report on Mandatory Reporting of MedicalNegligence. Sydney: NSW Parliament; 2000.

11. NSW Health Department. Qualified Privilege for QualityImprovement Committees and Programs in Health: Issues paper forcomment. Sydney: NSW Health Department; 2000 (publicationno. CRCP 00183).