Improving quality and safety of hospital care: a reappraisal and an agenda for clinically relevant reform


  • I. A. Scott,

    Corresponding author
    1. 1 Department of Internal Medicine and Clinical Epidemiology, Princess Alexandra Hospital, Brisbane, Queensland, Australia, 2Department of Medicine, University of Auckland, Auckland and 3Department of Medicine, Hutt Valley Health, Wellington, New Zealand
    Search for more papers by this author
  • 1 P. J. Poole,

    1. 1 Department of Internal Medicine and Clinical Epidemiology, Princess Alexandra Hospital, Brisbane, Queensland, Australia, 2Department of Medicine, University of Auckland, Auckland and 3Department of Medicine, Hutt Valley Health, Wellington, New Zealand
    Search for more papers by this author
  • and 2 S. Jayathissa 3

    1. 1 Department of Internal Medicine and Clinical Epidemiology, Princess Alexandra Hospital, Brisbane, Queensland, Australia, 2Department of Medicine, University of Auckland, Auckland and 3Department of Medicine, Hutt Valley Health, Wellington, New Zealand
    Search for more papers by this author

  • Funding: None

    Potential conflicts of interest: None

A/Prof Ian A. Scott, Department of Internal Medicine and Epidemiology, Level 5A, Princess Alexandra Hospital, Ipswich Road, Brisbane, Qld, 4102, Australia.


Improving quality and safety of hospital care is now firmly on the health-care agenda. Various agencies within different levels of government are pursuing initiatives targeting hospitals and health professionals that aim to identify, quantify and lessen medical error and suboptimal care. Although not denying the value of such ‘top–down’ initiatives, more attention may be needed towards ‘bottom–up’ reform led by practising physicians. This article discusses factors integral to delivery of safe, high-quality care grouped under six themes: clinical workforce, teamwork, patient participation in care decisions, indications for health-care interventions, clinical governance and information systems. Following this discussion, a 20-point action plan is proposed as an agenda for future reform capable of being led by physicians, together with some cautionary notes about relying too heavily on information technology, use of non-clinical quality personnel and quantitative evaluative approaches as primary strategies in improving quality.


Ten years ago, the Australian Quality in Healthcare Study showed 17% of all hospital admissions were associated with adverse events related to health care, the majority considered preventable.1 Two years later saw the establishment of the Australian Council of Safety and Quality in Healthcare (ACSQH) (renamed in 2006 as the Australian Commission on Safety and Quality in Healthcare), which, till 2005, had instigated more than 20 different programmes aimed at making care safer from which numerous strategies have been endorsed for widespread application by the Australian Health Ministers Council (AHMC).2 But has hospital care improved? Unfortunately, not as much as we would hope. More than half of 670 hospitals recently reviewed by the Australian Council of Healthcare Standards (ACHS) had inadequate quality-control systems for patient safety.3 US experience also laments slow progress towards safer care over the 5 years to 2005.4 Although much good work has been carried out, there is a sense that at the coalface of hospital medicine not a lot has changed. Why not? And what can be done about it?

This review undertaken on behalf of the Internal Medicine Society of Australia and New Zealand (IMSANZ) presents an analysis of underlying factors responsible for poor quality and unsafe care and then proceeds to offer a 20-point action plan for improvement, with the focus being on in-hospital care. To save repetition, when state or territory jurisdictions are quoted in the Australian context, district health board should be substituted for the New Zealand context.

Time to reconsider basic tenets

This paper contends that present quality and safety improvement programmes may be over-investing time and effort in non-core, second-order strategies which, although in themselves worthy, do not tackle the core or first-order clinically relevant issues central to making hospital care visibly safer. Many programmes comprise top–down, centralized, formal structures and processes (Table 1) which target occurrences that place the hospital, as an organization, at high legal risk, such as wrong-site surgery or egregious medication misadventure. However, although catastrophic and very preventable, these events are comparatively rare.5 Also, despite face validity, many of these safety initiatives, working in isolation, lack hard evidence of effectiveness in real-world settings.6–10 In contrast, common safety problems (e.g. nosocomial infection, contrast media-induced nephropathy and pressure ulcers) may be more amenable to an epidemiological approach involving surveillance, intervention and re-assessment and which adopts evidence-based strategies for improving patient safety (Table 2).11

Table 1.  Quality and safety procedures of unproven effectiveness and scope
Sentinel event reporting
Root cause analyses (of sentinel events)
Formal risk registers and risk management programmes
External hospital accreditation surveys and programmes
Quality/safety or clinical governance units
Open disclosure policies
Public reporting of quality and safety indicators
Table 2.  Evidence-based patient safety practices
  1. DVT, deep vein thrombosis; PTE, pulmonary thromboembolism.

Prevention of venous thromboembolism
 Identify patients at high risk of DVT/PTE and provide appropriate antithrombotic prophylaxis
Prevention of perioperative complications
 Use beta-blockers perioperatively in patients at high risk to reduce cardiovascular mortality
 Localize specific surgical procedures to high-volume centres
Prevention of nosocomial infection
 Implement and monitor protocols for decontaminating hands between patients
 Use maximum sterile barriers while placing central i.v. catheters
 Use real-time ultrasound guidance for central line insertion
 Use antibiotic-impregnated central venous catheters
 Institute appropriate use of antibiotic prophylaxis in surgical patients
 Institute continuous aspiration of subglottic secretions to prevent ventilator-associated pneumonia
 Vaccinate health-care workers against influenza
Prevention of pressure ulcers
 Use risk stratification procedures to identify patients at risk of developing pressure ulcers
 Use pressure-relieving bedding materials
 Evaluate patients for evidence of malnutrition and institute early enteral nutrition in critically ill and surgical patients
Prevention of contrast media-induced renal failure
 Use validated protocols to evaluate patients at risk for contrast media-induced renal failure and implement protocols for reducing risk of renal injury
Patient advocacy
 Ask patients (or legal surrogate) to recall and restate what they have been told during the informed consent process
 Inform patients undergoing high-risk elective surgical procedures of the likely reduced risk of an adverse outcome if carried out at high-volume facilities with superior expertise
 Document and publicize patient’s preference for life-sustaining care (including reference to advanced health directives)
 Implement patient self-management programmes for warfarin use in appropriately selected patients
Prevention of medication errors
 Actively involve clinical pharmacists in the medication use process: pharmacist liaison with prescribers on medication ordering and review (e.g. attendance of pharmacists on ward rounds), pharmacist-mediated medication reconciliation on admission and at discharge, pharmacist-mediated dispensing and monitoring of medications and patient education
 Use only standardized medication sheets, standardized abbreviations and dose designations and standardized labelling, packaging and storing
 Institute computerized prescribing systems that include alerts on adverse drug reactions
 Ensure medication storage and preparation areas are orderly, well lit, secure and clean; medications that look similar are stored separately from each other and high-alert drugs (e.g. i.v. electrolytes, adrenergic agents, chemotherapy drugs, neuromuscular blockers, insulin and narcotics) are clearly identified
 Dispense medications in unit-dose or, when appropriate, unit-of-use form whenever possible
Wrong-site, wrong-patient surgery
 Implement standardized protocols to prevent occurrence of wrong-site, wrong-patient procedures
Patient care information transfer
 Implement clinical handover procedures that ensure care information is transmitted in timely and comprehensible fashion to all who need to know
 Implement standardized systems for generating patient care summaries whenever patients move from one care team to another
 Specify explicit procedures for ensuring adequate levels of nursing staff based on usual casemix and experience and training of staff
 Ensure all patients in critical care units (intensive care units, coronary care units and high dependency areas) are managed by physicians with appropriate training and certification

Currently involvement of clinicians in safety enhancement tends to focus on repeated case finding, reporting and analysis of error and risk and adhering to safety procedures of unproven efficacy devised and imposed by non-clinical personnel. It could be argued that greater dividends in patient safety could be achieved if clinicians were more actively involved in developing, implementing and monitoring effective strategies for preventing frequently occurring problems that have already been well characterized.11 Resources and publicity may need to be redirected back towards basic determinants of high-quality hospital care which are discussed below under six themes.

Clinical workforce

The ‘Dr Death’ scandal in Bundaberg in Queensland centred on a fraudulent, incompetent surgeon undertaking complex operations simply because he filled a vacancy and earned the hospital income by carrying out high-paying elective surgery.12 The Davies report13 handed down in December 2005 identified: (i) weak penalties for practitioners misrepresenting themselves to registration boards; (ii) no requirement for non-clinical managers, in circumstances of their own choosing (such as ‘area of need’), to seek verification of a doctor’s skills from the relevant specialty college and (iii) managers dismissing expressed concerns about professional misconduct, citing insufficient performance data on which to take action. It is uncertain whether these circumstances are any better in other states and territories or for nurses and other health disciplines.

Inquiries at Campbelltown and Camden hospitals,14 King Edward Memorial Hospital15 and the Canberra Hospital16 all disclosed institutions trying to provide medical or surgical services with inadequate levels of staff, expertise or supervision and where these deficiencies were organizationally ignored.

Hospitals require adequate numbers of well-trained, well-supervised and properly credentialed health professionals skilled in safety and quality improvement, rostered appropriate hours, paid appropriately and professionally and legislatively empowered and protected to raise concerns and challenge authority and conventions in situations where they perceive potential danger to patients. Poor staffing levels,17–19 inadequate supervision,20 tightly constrained budgets and organizational cultures which suppress dissent21 characterize hospitals with disproportionately high rates of error and harm. The worsening shortage of skilled health professionals, especially nurses, coupled with changes to safer but shorter work shifts have important implications for patient care, which need urgent attention.22


Modern health care consists of care provided by doctors and nurses trained in more than 50 different types of medical specialty, as well as 10 streams of allied health, all operating in microsystems at a ward, unit or departmental level.23 Good teamwork with blending of roles, proper communication, close collaboration and sharing of authority is required if the more than 200 interprofessional interactions that occur on average during a single episode of patient care are to result in efficient, safe care.24 Although ‘we must work as a team’ is a frequently uttered phrase, the determinants of well-functioning teams (Table 3) are infrequently enunciated or promoted.25,26 This is a major failing as entrenched professional cultures and practice norms based on individual autonomy constitute daunting barriers to creating habits and beliefs of common purpose, accountability and interdependence that a safe culture requires.27 Hospitals characterized by dysfunctional or poorly coordinated teams with ingrained failures of interdisciplinary communication show excess mortality and morbidity.28

Table 3.  Attributes of well-functioning clinical teams
Common purpose and vision related to delivering high-quality patient care
Open and clear communication with transparency of processes
Participative leadership and balanced participation by all members
Defined roles and tasks with adequate levels of autonomy, skill, time and support to enable productive work
Mutual respect, collaboration and cooperation
Agreed methods of negotiation and conflict resolution
Effective decision-making based on objective data and opinion
Valued diversity with respect for differing professional roles and aspirations
Cohesiveness and regular meetings of all team members
Recognition and reward for individual contributions and group successes

Patient participation in care decisions

In making decisions about their care, some patients will be happy to defer completely to clinicians, others will ask to hear the options and understand rationales behind specific recommendations, whereas the remainder will actively seek out all relevant information and articulate their own preferences.29 In all three situations, however, better outcome and less harm can be achieved by using good clinician–patient communication skills, exploring patients’ attitudes and beliefs, sharing information on risks and benefits in ways patients understand and tailoring the way advice is provided to suit individual circumstances.30 Practical means for skilling patients in how they themselves can decrease their risk of iatrogenic adversity by way of education,30,31 decision aids32 and disease management approaches33 are listed in Table 4.

Table 4.  Patient empowerment strategies
  1. ACSQH, Australian Council of Safety and Quality in Health Care.

Orientation procedures that profile patients’ responsibilities in assisting caring clinicians and detail common safety risk situations and the options open to patients (and carers) for voicing their concerns about any aspect of care. Such procedures should feature named photographs and contact details of key members of the treating team.
Educational resources on how patients can play a more effective role in managing their own illness, improving safe medicine use, avoiding needless interventions and optimizing psychosocial support. Examples include brochures produced by ACSQH –10 Tips for Safer Health Care (available at – and US Agency for Healthcare Research and Quality –Patient Fact Sheet. 20 Tips to Help Prevent Medical Errors (available at
Regular briefing of the patient on provisional diagnosis(es), likely clinical course and prognosis, intended management (investigations, treatments, consultations and invasive procedures), expected date of discharge and timing of consultations with carers and family members
Promotion of active patient participation in care decisions by use of evidence-based patient decision aids in situations involving narrow treatment benefit–risk ratios; advance health directives, advance care planning and elicitation of preferences for end-of-life care and patient support groups (including those based on ethnicity or religion). An excellent National Health and Medical Research Council of Australia resource is Making Decisions about Tests and Treatments. Principles for Better Communication between Healthcare Consumers and Healthcare Professionals (available at
Predischarge kits that contain disease/diagnosis information sheets, management plans and treatment goals, medication lists, self-management resources (including contact details of patient support groups) and follow-up schedules
For elective encounters, pre-admission and pre-clinic orientation kits which outline the date, nature and reason for the encounter; any clinical material patients should bring with them (copies of past investigation results, specialist letters, discharge summaries from other hospitals, lists of current medications (including over-the-counter drugs) and details about allergies or past adverse events); names and contact details of their general practitioners and other treating specialists and a listing of the problems (symptoms and disabilities) of greatest concern to them and what they would like to see achieved by the end of the consultation/admission

Indications for health-care interventions

Three major threats to patient safety arise from the use of diagnostic and therapeutic health-care interventions: (i) overuse (receiving care of little or no value), (ii) underuse (failing to receive needed care) and (iii) misuse (errors and defects in delivering care).34 Where it leads directly to physical injury, misuse of technology – a failure of execution – is the risk that most captures public (and media) attention and incites emotional reaction. Yet it is probable that the more indirect (and masked) burdens of complications and harm, lost productivity and increased costs resulting from overuse or underuse of care – problems of decision quality – are orders of magnitude greater than those resulting from misuse.

Up to 30% of clinical investigations and treatments confer little or no gain in patient health,35 with little potential for benefit for many patients undergoing coronary revsacularisation,36 arthroscopy,37 caesarean section,38 breast cancer screening,39 prostate surgery40 and carotid endarterectomy.41 Potentially dangerous medications such as non-steroidal anti-inflammatory drugs are prescribed to thousands of patients.42 Conversely, studies show consistent failure in providing effective treatments in prevalent, life-threatening conditions such as acute coronary syndromes,43 heart failure,43,44 stroke,45,46 diabetes44 and kidney failure.45 US data suggest that only 50% of patients receive recommended preventive care with regards to common causes of major illness.47

In light of the above, a predominant focus on catastrophic but infrequent harm and ‘near misses’ that are directly attributable to errors of task execution ignores a much greater reservoir of potential iatrogenesis. Narrowing gaps between best practice and routine care by deploying effective systems of clinical decision support, practice evaluation and care improvement (Table 5) may realize much larger dividends in improved patient outcomes than relying solely on administrative instruments such as sentinel event reporting or risk registers.48–52

Table 5.  Attributes of effective clinical decision support, practice evaluation and quality improvement
Clinical decision support (guidelines, protocols, pathways, reminders and prompts)
 Evidence based, credible and locally agreed
 Readily accessible and usable at the time and locale of decision-making
 Automatically provided as part of clinician workflow
 Delivers practical and feasible recommendations for key care processes graded according to strength of evidence, consensus of opinion and applicability to patients
 Openly endorsed and promoted by senior clinicians and opinion leaders
Practice evaluation (audit and feedback, clinical indicators and process measures)
 Accurate and clinically credible clinical management and outcome data relevant to well-defined clinical conditions or patient groups
 Timely feedback of relevant, readily assimilable information to all clinicians involved in care
 Evaluation undertaken at level of clinical microsystems (unit and department)
 Comparison of current practice standards with best practice and peer group standards
 Commitment to enact remedial strategies in response to evaluation feedback
Quality improvement (various strategies for improving clinical practice)
 Detailed analysis of current care processes (process mapping) that defines reasons for suboptimal care identified during evaluation phase
 Formulation and implementation of quality improvement strategies as a collaborative effort involving all key clinicians
 Incremental plan-do-study-act approach to practice improvement that starts with problem awareness and uses enabling and reinforcement strategies which enhance clinician ability to put knowledge into practice
 Practice improvement strategies chosen on the basis of empirical evidence of effectiveness (e.g. reminders, prompts, computerized decision support, audit and feedback), buy-in from all key stakeholders and support from senior management

Over and above care related to specific clinical conditions is the uneven application across institutions of a number of effective, evidence-based safety practices aimed at preventing common hospital-acquired complications (Table 2).11,53,54 In addition, redesigning high-risk clinical processes based on evidence obtained from high reliability/high-hazard industries such as aviation can substantially improve safety, wherein reliance on individual vigilance (paying attention) for avoiding mistakes is replaced by systems with specific ‘forcing functions’ (such as unambiguous feedback, displays, instructions and automated systems).55,56

Clinical governance

Every hospital is expected to have formal surveillance and accountability procedures or ‘clinical governance’ framework for improving quality and safety despite uncertainty as to their effectiveness.57 To succeed, these frameworks must (i) devolve governance and performance monitoring to the level of clinical units or departments where care is provided and whose staff carry ultimate responsibility;58 (ii) feature practice-relevant, data-driven agendas that actively seek involvement and innovation on the part of practising clinicians and (iii) require hospital executives, clinical governance units and quality improvement coordinators to sponsor and support quality and safety activities within departments (or across departments in situations of shared care).

These activities may, at times, require authorization of additional funding for performance measurement, enhancement of service capacity or redesign of care delivery to address serious problems in quality and safety that have been clearly identified or are strongly suspected. At present, hospital funding formulae are based on activity and throughput rather, as in some countries, on patient outcomes linked to quality and safety under ‘pay-for-quality’ principles.59,60

Hospital accreditation procedures of the ACHS, although becoming more stringent, do not mandate hospitals to implement proven safety practices or mandate payers (government health departments or private health funds) to withhold funding to hospitals that consistently fail to implement such practices.3 None of the previously mentioned hospital inquiries was circumvented despite each of the relevant hospitals having recently received ACHS accreditation and despite attempts by ACSQH to improve sentinel event reporting and clinical governance.

Instead, it was individual whistleblowers who, at great cost to themselves, finally raised enough public pressure to force authorities to act in the face of compelling anecdotal evidence.21 In each case, organizations chose to respond to information indicative of safety problems by suppressing it or manipulating it to maintain the status quo (pathological response) or by constructing new administrative procedures which failed to address underlying causes (bureaucratic response). Instead, a generative approach would have used such information positively for detailed inquiry and fundamental system change.61

Information systems for performance monitoring and safety enhancement

Clinical databases and audit systems that allow prospective collecting and reporting of clinical performance data remain severely underdeveloped.62,63 Constant peer review and self-review informed by such data may identify and remedy poor performance more effectively than delayed, ad hoc retrospective sentinel reporting and chart review.64,65 Routine administrative data, traditionally used to measure output efficiency (length of stay, costs and utilization rates), could, using appropriately validated techniques, be transformed into useful, real-time quality and safety measures.66 Linkage of pharmacy, laboratory and patient databases could flag drug–result and drug–patient scenarios with potential for adverse events.67 Systematic individual-based linkage of hospital-episode statistics, death registries and clinical databases would allow longitudinal relations between process and outcomes of care to be more easily ascertained.68,69 Multihospital, data-driven collaborations aimed at improving care in specific clinical areas have potential to shift routine care significantly towards best practice standards in relatively short timeframes.43,70,71

Clinically relevant action plan for improving quality and safety

The following actions might, given adequate professional, managerial and political will, help to significantly improve current levels of quality and safety of care.

Action 1: Simplify external professional registration

Establish sets of standardized procedures for registration for each discipline of health professionals at the national level under the jurisdiction of disciplinary boards or authorities affiliated with the relevant college or academy. These boards, which might be housed within one central agency, would collate and verify relevant information concerning each registration application received from specialty colleges, previous employers and state- or territory-based registration authorities. Harsh penalties would apply to what are regarded as criminal offences of practitioner non-adherence or fraudulent misrepresentation. ‘Area of need’ exemptions to standard procedures that involve less rigorous vetting of professional competencies should be abolished.

Action 2: Implement institutional professional credentialing

As a condition of employment or visiting rights at any hospital, all practitioners should (i) have their credentials cross-checked with data stored in the central agency, (ii) agree to a transparent and periodic process of peer-mediated performance appraisal and recredentialing, (iii) commit to continuous monitoring of processes and outcomes of their practice using indicators appropriate to the discipline and level of expertise and (iv) participate in a relevant college-ordained maintenance of professional standards programme.72

Action 3: Augment external accreditation procedures

External hospital accreditation surveys should include not only verification of hospital-wide quality and safety procedures but also direct sighting of departmental clinical and safety profiles, professional appraisal data and diagnosis-specific care process and outcome measures if hospitals are to be granted unconditional accreditation.

Action 4: Define institutional service capacity

Each hospital and its respective departments should define their service capacity using a transparent and data-driven process, that is, the clinical conditions and patient types for which each has the expertise, facilities and staff to manage well, and the circumstances under which particular patients should be referred or transferred to another team or institution. Service capacity should be defined for both normal operating conditions and temporary situations of serious capacity–demand mismatching wherein surges in patient load or shortfalls in staff or resources exceed predetermined safe care thresholds.

Action 5: Implement institutional governance procedures

State government executives in liaison with district and health area managers should formulate an action-based clinical governance framework that applies to all hospitals and is linked to a timeline for its enactment and a commitment to provide required resources for achieving its objectives. Although the scope and number of specific governance actions will differ between institutions, a basic, generic framework should apply to all.

Action 6: Enact whistleblower protection

In cases of professional incompetence that may require earlier-than-scheduled review of credentialing, there needs to be legal and vocational protection for clinician ‘whistleblowers’ voicing legitimate concerns about the competency of specific individuals. There needs to be a process whereby whistleblowers are given access to professional appraisal panels empowered to independently hear and investigate such concerns free of political, legal, managerial or media interference. More generally, staff at all levels should feel encouraged to voice concerns about organizational, managerial and industrial issues that appear to be impacting adversely on patient safety.

Action 7: Provide professional training in quality and safety

Health departments, unions, university health faculty, specialty colleges and medical societies should collaborate in forging new training programmes and career paths that focus on optimizing quality and safety.73,74 A key reform is to enable doctors, nurses and allied health professionals to more effectively undertake high-order bedside tasks that require specialized training by devolving distracting, low-order clerical and non-clinical procedural tasks to appropriately trained assistants.75 In times of future shortages of specific groups of skilled clinicians, professional boundaries (and associated training processes) will need some renegotiation in situations where evidence confirms that tasks previously considered as the exclusive province of one type of practitioner can be safely carried out by other groups.75

Action 8: Promote multidisciplinary teamwork

All hospitals should formalize and support multidisciplinary team activities that have been shown to lower patient risk of harm and promote interdisciplinary programmes of education and training in safety procedures, team skills and communication. Affiliated university schools of medicine, nursing and allied health should implement interdisciplinary activities that inculcate attitudes of respect and cooperation among members of different professional streams.

Action 9: Establish clinical orientation procedures

All hospital departments should provide all newly appointed clinicians with a multidisciplinary orientation and information package that outlines departmental policies and procedures and personnel roles and responsibilities.

Action 10: Formalize clinical handover and care coordination procedures

All hospital departments should have explicit procedures for (i) clinical handover between shifts for all disciplines and between different care teams;76 (ii) accessing information and advice from external caregivers (such as general practitioners and private specialists) with regards to inpatient care; (iii) recording clinical information arising from ward rounds, case conferences and consultant visits and conveying it to others who need to know; (iv) expeditious identification and management of patients who become clinically unstable;77 (v) use of paging, phone and other call systems that minimize potentially unsafe disruption of clinical work and allow prioritization of tasks78 and (vi) transferring structured patient information (discharge summaries, clinic letters and community referrals) to general practitioners and other clinicians involved in post-hospital care.79

Action 11: Enact clinical governance at the work unit level

All hospital departments should hold regular, minuted multidisciplinary business meetings (including representation from non-clinical staff) in which quality and safety issues are standing agenda items, where unit performance indicators are periodically tabled and discussed and where innovative ideas and projects for improving care are actively encouraged.

Action 12: Implement practice improvement at the work unit level

All hospital departments should develop an ongoing quality and safety programme centred on one or more departmental personnel acting as patient quality and safety officers with defined roles and protected time. Such personnel could act to locate, adapt and promulgate practice improvement interventions that have proven successful in other settings, and to liaise with multisite collaborations that focus on common problems.

Action 13: Provide patient-centred care information

All hospital departments should provide counselling and written materials to patients on admission and at discharge that address the issues listed in Table 4.

Action 14: Formalize patient advocacy

All hospital departments should assign the role of ‘patient advocate’ to a patient representative whose role is to discuss with patients (i) their rights and responsibilities in receiving care, (ii) any safety and quality concerns they have relating to their care and (iii) circumstances surrounding any error or mishap that may occur.

Action 15: Pilot unique patient identifiers and patient-held information systems

National government should consider piloting a nationwide unique patient identifier system, coupled with an opt-in, patient-held ’smart card’ health record. These devices would allow patient health information, necessary for safe and seamless care, to be available at any time in any hospital. The huge benefits of reduced test duplication, less medication errors and timely access to medical history (especially in circumstances where patients are rendered unable to provide such information) will likely far outweigh any downside from potential breaches of privacy.

Action 16: Implement evidence-based decision support systems

All hospital departments should develop and implement readily accessible, evidence-based clinical practice guidelines, protocols and other forms of decision guidance covering their most commonly encountered conditions, procedures and treatments. All staff should be made aware of the availability of such guidance and be trained in its use. Each department (or group of departments) might appoint a ‘clinical informationist’ or ‘knowledge translator’ with clinical training and skills in evidence retrieval and appraisal who would steer the formulation of guidance with input from senior clinicians.80 The process of formulating such guidance might identify practices associated with marginal benefit or potential safety concerns that would warrant formal departmental disendorsement.

Action 17: Implement evidence-based patient safety practices

All hospitals should attempt implementation of patient safety practices listed in Table 2 across the institution using methods suggested by others as being effective, closely monitor compliance and apply sanctions to departments that consistently fail to meet agreed standards.11

Action 18: Standardize and simplify high-risk care technologies

All hospitals should simplify and standardize ‘high patient risk’ technical procedures throughout the organization such as fluid and medication prescribing and giving, particularly in relation to ‘high-risk’ regimens; use of infusion pumps, anaesthetic machines and other medical equipment and reporting and review procedures for significantly abnormal investigation results.81 Allowing different departments to retain separate systems on grounds of historical tradition or idiosyncratic preference, in the absence of supporting evidence of patient benefit, should be phased out.

Action 19: Implement data-driven performance monitoring systems

All hospital departments should measure, analyse and report a set of specialty-specific process-of-care and outcome clinical indicators and a set of hospital-wide (or generic) safety indicators. Clinical indicators would ideally relate to high-volume or high-risk conditions/procedures,82–84 whereas generic safety indicators might include those listed in Table 6.85

Table 6.  Generic patient safety indicators
  1. DVT, deep vein thrombosis; PTE, pulmonary thromboembolism.

Unexpected deaths
Unplanned transfers from ward to intensive care unit or high dependency unit
Unplanned returns to operating theatre or catheter laboratory within 48 h of procedure
Unexpected and serious periprocedural complications (e.g. cardiorespiratory arrest and hypovolaemic shock)
Serious medication or fluid errors
Serious nosocomial complications: infection, DVT/PTE, advanced pressure ulcer, falls resulting in fracture or major soft-tissue injury
Unplanned re-presentations within 7 days of discharge
Adverse findings from coroner inquests or autopsy reports

Although indicators might be analysed in block format (i.e. at the end of a prespecified reporting period for a unit or department), ‘real-time’ continuous monitoring using statistical process methods (i.e. run and control charts86,87 or cusum analysis88) is likely to be more informative for infrequent occurrences such as deaths and sentinel events and can be potentially applied at the level of individual clinicians for high-volume, high-risk procedures.

Primary data sources for either type of indicator may include, depending on the indicator, administrative databases, hospital records, discharge summaries, mortality reviews, coroners’ reports and patient complaints. Indicators for each department (but not necessarily those pertaining to individual clinicians) should be accessible to anyone who can substantiate a legitimate interest in reviewing them.

Action 20: Apply penalties for consistently poor performance

Hospital management (or district health boards or, in the case of private hospitals in Australia, the Health Insurance Commission and private health insurance funds) should consider applying financial and professional sanctions to departments or clinician groups that show consistently unacceptable rates of preventable errors or complications resulting from poor governance procedures or continuance of unsafe clinical practices.89

Some cautions

Although electronic health records, computerized ordering and prescribing systems and clinical databases hold much promise, information technology (IT) is often oversold as a definitive quality and safety solution. In improving care, we need to closely analyse human factors – how humans interact with each other and with their environment – and not place undue reliance on putative advances in IT, especially when novel but poorly designed IT systems may in fact add new safety problems.90,91

In addition, clinical governance ‘roadmaps’ designed and overseen by quality officers and senior management divorced from those directly involved in clinical work are doomed to fail. Front-line staff who deliver care must be allowed to help design and implement systems of governance and innovation and ensure an appropriate blend of feasible checks and balances.23,58

Finally, and most importantly, in improving quality and safety, we must confront and deal with the chaotic unpredictability of hospital practice, which continues rapidly to evolve and grow in complexity. Interactions between clinicians and their environment are those of ‘complex adaptive systems’ in which the how, why and what of clinical processes are in constant flux.92 Consequently, if we are to propose feasible and sustainable methods for improving care, we will need to blend qualitative approaches (based on cognitive psychology, social learning theory, systems analysis and political science), which attempt to ascertain why humans do what they do, with quantitative approaches (evidence-based medicine and biostatistics), which suggest what things may make care safer and of higher quality.93–95 Blending these two approaches will be essential if we are to successfully renegotiate traditional mindsets around clinician autonomy, craftsmanship models of professionalism and managerial leadership in ways that are acceptable to all players.96,97

Next steps

An increasingly important role of physicians, and of our representative organizations such as IMSANZ and the Royal Australasian College of Physicians, is to provide clinical leadership in promoting an effective quality and safety agenda. In their recent review of patient safety in the USA, Leape and Berwick recommended a meeting of all key stakeholders in health care with the aim of formulating and implementing a national action plan that attracted commitment from every major player.4 Perhaps the time has come for the ACSQH and its New Zealand counterpart to convene national summits with similar objectives, which send a clear message to government and health funders that physicians together with other health professionals are committed to improving the safety and quality of hospital care in Australia and New Zealand.


The authors thank Peter Greenberg, Bruce King and all current IMSANZ councillors for their review and comments relating to the IMSANZ position paper on quality and safety from which this paper originated.