Sharing the pain: lessons from missed opportunities for healthcare improvement from patient complaints and litigation in the Australian health system

The patient safety movement, a response to the recognition that patient harm was a common event, is nearly 50 years old.¹ It is also now some 25 years since the publication of a landmark analysis of adverse incidents in hospital.² In that report, nearly 4% of all hospital admissions were complicated by at least one episode of avoidable harm. It is confronting that, more than 25 years later, the rate of adverse events in Australian public hospitals is approaching 7%,³ and medical error remains the third leading cause of death in hospital.⁴ Equally challenging, our healthcare systems continue to be plagued by whole-of-service quality and safety failures, such as occurred in King Edward Memorial Hospital in Western Australia (2000),⁵ Bundaberg Hospital in Queensland (2005),^6,7 Djerriwarrh Health Services in Victoria (2014),^8,9 Bankstown–Lidcombe in New South Wales (2016)¹⁰ and, most recently, Oakden in South Australia (2017).¹¹ Whole-of-service failures often result in a government-commissioned review, such as those conducted under the auspices of the UK’s National Health Service (NHS) on Bristol (2001),¹² Mid Staffordshire (2013)¹³ and Morecombe Bay (2015)¹⁴ or, as was the case in Australia, with the Queensland Public Hospitals Commission of Inquiry following the Bundaberg hospital events⁶ and Targeting Zero,⁸ the review into the Victorian public hospital system following 11 potentially preventable perinatal deaths at Djerriwarrh Health Service. Common to these reviews has been the call for improved information sharing and shared learning to prevent such failures in the future. Indeed, the Duckett review explicitly called for the establishment of an agency (Safer Care Victoria) within the Victorian Department of Health and Human Services to facilitate information sharing and timelier shared learning.⁸ Although some progress towards better information sharing has been made in many Australian jurisdictions, we believe that there are other, as yet untapped, opportunities to learn from harm. Specifically, the analysis and shared reporting of individual patient complaints and medicolegal claims may offer the potential to enhance patient safety in a more strategic manner. Following significant adverse events, formal complaints and medical litigation are the two main avenues through which the healthcare system is held accountable and, when relevant, patients are recompensed for their losses.¹⁵ These events generate potentially insightful data on care provision where there has been either a real or perceived deficiency. As such, these events may afford the opportunity to learn from past mistakes and thereby systematically direct health service and system improvement using the patient voice in a manner not afforded by other approaches.

Indeed, with the aim of improving the understanding of medicolegal claims and complaints, researchers have looked at both practitioner and craft group factors that affect the distribution of medicolegal events. One interesting feature of these events is that their distribution at the doctor level is not equal.^15–17 Medicolegal complaints and claims commonly cluster among a small minority of practitioners.^16,18 Curiously, there has been very little research into how the patient and care factors that contribute to medicolegal events may be used for learning. Our current understanding of adverse medical events resulting in complaints and medicolegal claims is sufficient only to determine that they concentrate to a minority of doctors and key speciality groups.^16–18 Many health service failures, including Bundaberg and Djerriwarrh, have issues related to practitioner competence and communication in common.^7,8 That these factors continue to be regularly identified in root cause analyses of healthcare incidents highlight that our ability to learn from failure, prevent adverse outcomes and protect future patients remains limited.^6,8,19 There is a need to better understand the reasons behind the system’s inability to learn from adverse events and how this may be changed. In that regard, herein we review the various actors in the patient complaint and medicolegal claims frameworks in Australia. We hope to highlight one approach by which we may better learn from these events, consider where such opportunities may lie and suggest how improved system-wide information sharing, as was recommended for sentinel events,⁸ may be established using existing data collections with limited additional cost. Although we draw on the Victorian landscape and, to a lesser degree, New South Wales and Queensland, we hope that there are sufficient similarities across all Australian health jurisdictions to allow broader extrapolations. Ultimately, our aim is to explore how we may further help the Victorian healthcare system, and those of other Australian jurisdictions, to improve the use of learnings from non-traditional patient safety systems, like patient complaints and litigation.

Patient complaints about their health care are a rich source of quality improvement intelligence that is relatively unmined and less than adequately shared. There are several avenues through which a patient may direct a complaint about their care, and the processes for handling health service complaints in Australia vary slightly from state to state. Complaints may be directed to individual practitioners (a common route for private patients), the Health Complaints Commissions (HCCs), the Australian Health Practitioners Regulation Agency (AHPRA) and its associated boards or to government departments of health. However, at least in the first instance, complaints regarding health care are most commonly addressed directly to the health service concerned. All health services have a complaints or patient liaison officer (or equivalent) who is responsible for the receipt and coordination of patient complaints. This representative would typically receive all patient or family complaints, either directly or indirectly depending on who the complaint was made to. The responsibility to coordinate the response to the complaint in consort with hospital management, clinicians, the legal team and others as required typically falls under this role. These direct-to-service complaints are investigated and typically resolved internally. It would be unusual for monetary compensation to be awarded through this mechanism. Patients using this avenue of communication are usually seeking an acknowledgement and understanding of their experience, an apology and, not infrequently, some sort of system change to reduce the likelihood of a future patient having a similar experience. In this latter regard, although hospitals may introduce changes that improve care and reduce the future likelihood of adverse events, these changes are often only at an individual unit level, with little sharing across the hospital. Further, there are no mechanisms to routinely share changes at a system level, across hospitals. As such current hospital-based complaint management systems are potentially missing key opportunities for system-wide improvement.

Presently, individual health services have the opportunity to collate their complaints experiences to allow for scheduled and regular reviews of the common complaint themes and use subsequent recommendations to inform improvement across their whole service. Such collation and thematic interrogation is neither a common nor uniform feature. Indeed, mishandling of patient complaints was highlighted as one of the missed warning signs in the review into the Mid Staffordshire health service failure in the UK^13,20 and was a common feature of the poorest performing NHS hospitals in England.²¹ In an Australian context, in hindsight, the complaints data at Bundaberg Base Hospital could have alerted management to the performance issues of Dr Jayant Patel.^6,8 Further, a lack of awareness and adequate response to patient complaints may also have contributed to the lack of timely recognition of the maternity events at Djerriwarrh Health Service in Victoria.⁸ Patient complaints are recognised as a useful source of data on patient risk.¹⁵ Indeed, the sharing of complaint data between AHPRA, the Health Complaints Commissioner and the Mental Health Services Commissioner was highlighted in Targeting Zero to be a key improvement priority for Victoria.⁸ However, we believe that this recommendation misses other potential opportunities for shared learning. Compared with complaints to individual health services, the complaints reported to AHPRA and the Commissioners are few in number. We suggest that it would be useful to have a mechanism for all health services to share patient complaint data, even when those complaints were only reported locally and not to any of the statutory authorities. We believe that such shared learning could afford opportunities for individual health service experiences to improve care at a whole-of-system level within a jurisdiction, such as Victoria. The infrastructure to support such collaboration already exists in Victoria. All health services use Riskman (Datix Ltd, London, UK), the third-party software currently licenced by the Victorian Department of Health and Human Services to record patient complaints and clinical incidents.^22,23 System-wide monitoring and reporting back to services of complaint frequency, as done by the Queensland Government’s Patient Safety and Quality Improvement Service,²⁴ would afford interhospital benchmarking, trends analyses and the potential for an earlier warning of a service in trouble. But more than that, a shared complaints system with a uniform taxonomy, such as has been reported recently,^25,26 would allow an analysis of complaints beyond simple rates. It would allow a thematic analysis at both an individual service level and at a system level. Such a simple program, with central coordination, has the potential to efficiently determine key quality shortfalls that could then be used target improvements. Of course, the eventual usefulness of this approach will depend on the clarity of patient complaints, how accurately complaints are documented in the central database and how effectively those complaints represent the key weaknesses in a service or a sector. At present, we have no evidence to show that such an approach to sharing complaints experiences and learnings can improve care at a system level because no jurisdiction has yet done this in a manner to target system improvement. The question we pose is whether it is worth exploring. We believe that it is. Given existing infrastructure, we do not believe that it would be overly expensive or burdensome to do so, at least in Victoria and probably in other jurisdictions.

In addition to using individual hospital complaints data, the HCCs were highlighted in Targeting Zero as a rich source of patient risk data.⁸ In Australia, these organisations are independent, state government-funded and state-specific bodies that focus on complaints arising from patient and/or family dissatisfaction with care,¹⁵ often, but not exclusively, related to adverse outcomes or patient harm. The HCCs principally provide mediation and conciliation services, assisting aggrieved patients to meet with health service clinicians to discuss their complaint(s) and better understand what happened. HCCs have no power to compel monetary settlement, but they do facilitate a settlement agreement between parties.¹⁵ Importantly, HCC findings and deliberations, including the outcomes of independent experts commissioned by the HCC to inform proceedings, are legally protected by qualified privilege and cannot be admitted as evidence in any subsequent litigation.¹⁵ All Australian state HCCs publish annual reports. Within their respective annual reports, each Commission provides a breakdown of complaints handled. All six state HCC reports also address health system quality improvement as an outcome. Indeed, many of the annual reports detail specific examples of how patient complaints have resulted in specific service improvements or practice changes. These data, where sufficiently granular, offer great promise in guiding interventions to improve health care safety and quality and the Commissions are overtly enthusiastic that this will be so. Indeed, the Western Australian Health and Disability Services Complaints Office lists system improvement implementation as a key performance indicator, and the first focus of its strategic plan.²⁷ The Victorian Health Complaints Commission lists it as a ‘major priority’.²⁸

There is evidence that purposeful improvement directed by patient complaints works. An individual practitioner’s history of complaints is predictive of future complaints against them.^29–31 A ‘dose–response’ relationship exists whereby each complaint a practitioner receives increases the likelihood of a subsequent complaint.^15,18 A greater number of patient complaints is associated with increased likelihood of medicolegal claims (a proxy measure for likelihood of serious preventable complication).¹⁸ Identifying problem practitioners to target reflection and retraining or sanctioning is core business for the healthcare safety and quality machinery of a jurisdiction. Such an approach has been used effectively at Vanderbilt University Medical School (Nashville, TN, USA) through their Patient Advocacy Reporting System (PARS).^32,33 PARS seeks to compile data on complaints against practitioners for indexation as a measure of medicolegal claim risk.^33,34 These practitioners are then targeted to receive ‘awareness feedback’, a process of peer review by specially trained practitioners.^32,33 Following this intervention, complaints are monitored as a specific improvement measure.³² The Vanderbilt group was able to demonstrate a marked improvement in practitioner complaint rates, and an associated significant decrease in professional liability claims.³² Where there is no improvement, or if there is further deterioration in complaint frequency following intervention, the individual practitioner is reviewed for ‘authority intervention’ (ranging from voluntary relocation to limitation of privileges or dismissal).³² Despite this experience, the use of patient complaints as a systematic component of quality improvement remains relatively underdeveloped in Australia.

A complaint about care may also be made to AHPRA. AHPRA is the body responsible for enacting the National Registration and Accreditation Scheme.³⁵ It is the peak body under which the 14 health profession boards, known as the National Boards, exist to regulate each professional stream.³⁶ AHPRA collaborates with each of the National Boards to fulfil both regulatory and investigatory functions. In contrast with complaints to either a health service or HCC, AHPRA notifications pertain only to complaints about an individual registered health practitioner and not about a service overall. Notifications relating to doctors where there is concern over a clinician’s professional conduct, competence or health affecting their clinical practice are referred to the Medical Board of Australia (MBA) for investigation.¹⁵ Where a finding of impaired practice or care is sustained, resolution methods used by the MBA may include re-education and further training, disciplinary charges and/or sanctions relating to a practitioner’s medical licence.^15,37,38 Despite these safeguards, inadequate investigation and oversight of medical practitioners continues to be implicated as an important factor in health service safety breaches.^6–8 One of the most notable examples of this in Australia was witnessed in the safety failings of Bundaberg Hospital, where the lack of competent care by a single doctor was implicated in the deaths of 13 patients and significant morbidity to many more.⁶ Despite this incident, health service failures sharing similar themes of communication, poor reporting culture and a lack of practitioner oversight have continued to occur in Australia.⁸ Such failures highlight the need for the collation and sharing of information across health services, health systems and health jurisdictions. However, in Targeting Zero, Duckett and his review team identified two barriers limiting the sharing of AHPRA findings.⁸ First, the organisation’s investigation process was both too slow and too complex to afford timely learnings. Second, the ‘right of a practitioner to remain anonymous throughout the whole reporting, investigating and decision-making process’ hindered data sharing.⁸ Instead, Duckett commented that the ‘priority must be to protect patient safety and the public interest’.⁸ Recently, AHPRA has faced similar criticism in the media for its protracted investigation process and lack of transparency.^39–42 Complainants are not provided with information about the status of an investigation until the investigation is complete and findings are made. This can take years from the time of the initial complaint. In addition, where allegations are serious enough for a health practitioner to be stood down from their employment at a public hospital, they may still be able to practice in a private capacity until an AHPRA investigation is finalised.^8,40 Such was the recent case of a neurologist who was reported to AHPRA for sexual misconduct with a young adult male patient.^40,43 These circumstances arise from AHPRA or, as was the case with Bundaberg, the state medical board being constrained by privacy legislation (see Privacy Act 1988 (Aus)) that favours practitioner confidentiality over patient safety.⁴⁴

Calls, like Duckett’s, for an integrated framework in Australia to collate and distribute learnings from health incidents are not new. They have been made since 2006.⁴⁵ Sadly, since then relatively little, if any, progress has been made towards the development of such a resource despite multiple health service failures. It is time to ask how many more incidents resulting in avoidable patient harm must the Australian public endure before a more coordinated approach to national sharing of healthcare complaints between agencies and across jurisdictions is created?

To be optimally effective, complaints data will need to be drawn and collated from several sources. This is of particular importance in jurisdictions such as Victoria where many medical practitioners have sessional appointments at several different hospitals, both public and private. Either the HCC or a central government quality and safety agency, such as Safer Care Victoria, are well placed to establish formal links with health services (public and private) to deliver a system-wide complaints analysis and reporting function. Currently, such a central multiagency approach to complaints to inform quality improvement is not operational in Australia. Instead, where it exists at all, quality improvement driven by patient complaints review is largely restricted to individual health services and is typically initiated from the review of a single case with learnings applied within an individual unit or service. The opportunity for trends analyses, benchmarking, preventative shared learnings across services and, of course, early identification of the failing practitioner or service remain unrealised. If a centralised resource with the capability to pool complaints information from organisations such as individual health services, AHPRA, the HCCs and government departments could be established, it is possible that safety and quality gaps would be identified with far greater speed and efficiency. It would certainly be worth exploring the opportunity.

In addition to the various complaints handling agencies, the ultimate avenue of redress for patients following an adverse event is through the civil judicial system. Although complaints and feedback may result in a negotiation for compensation occurring between parties, the courts remain the only means of compelling financial compensation following an act of negligence. As a breach of tort law, law related to civil wrongs and personal injury,^31,46 medical negligence in Australia is heard as a civil common law matter, except where there is evidence of criminal conduct.^15,47–49 The civil courts resolve cases through determination of negligence and, if sustained, subsequently by the assignment of monetary damages and costs.¹⁵ Although this financial compensation is an important facility for the individual, no formal process currently exists in Australia whereby the courts can inform healthcare reform to prevent similar future failures. Indeed, it is not a responsibility of the courts to have a role in contributing to patient safety. Nor is there currently a mechanism by which findings from litigation, whether settled before or through court proceedings, can be easily collated, analysed and shared at a system level to inform sector-wide improvements towards reducing future avoidable harm. The data generated by the courts through the civil litigation process may present another opportunity to learn. The value of settled tort claims as a source of quality data on patient safety has been recognised by experts internationally.^50,51 However, the process for capturing these data remains underdeveloped in Australia. So, although the legal framework to manage complaints and compensation subsequent to adverse events are well established, currently structured systems are limited in their capacity to reform future practice. Typically, the feedback of findings from medicolegal claims is conducted on an ad hoc basis, case by case. Recently, some law firms have started providing ‘lessons from losses’ summaries to the health service that engaged them in managing the claim. At present this process is principally driven by the insurer looking to prevent similar incidents occurring within the same health service or by like practitioners, thereby reducing liability over time. However, whether these lessons could be shared and used to inform other services to avoid future similar events has not been reported. Indeed, these summaries are not commonly shared beyond the parties directly involved in the case. Many private medical defence organisations (MDOs) also publish information on claim trends and useful learnings from individual cases as education for their membership. However, typically these reports are provided to the organisation’s membership and are not generally available for whole-of-system learning.

A key limitation to the potential use of these ‘lessons from losses’ and MDO educational materials is that, typically, claims are settled or litigated years after the incident. This means that lessons learned are often either redundant because practice has changed or they simply have less traction with the workforce due to the passage of time. In contrast with sharing patient complaints, it is difficult to see how lessons from litigation processes, such as sharing expert opinions, could be shared in a timely manner without prejudicing trial outcomes.

Nonetheless, there are several good examples internationally of using closed medicolegal claims to strategically improve quality of care and patient safety.^52,53 Despite this, in Australia, the legal profession is largely excluded from the process of safety and quality assurance. This is surprising given that medical malpractice is obviously intimately linked to patient safety.⁵⁴ Medical malpractice lawyers are ideally placed to provide valuable insight into the various factors that contribute to healthcare-associated harm. We believe that the limited use and dissemination of this rich cache of information on past mistakes is a significant missed opportunity for improving patient safety.

The use of incident factors as a component of medicolegal risk analysis is another method by which health services could gain insight into their quality and safety shortcomings. Medicolegal claims are typically categorised by the severity of injury.⁵⁵ The more serious the injury and long-term sequelae, the larger the payout.^29,55 Beyond this, not much is reported about the correlation between claim settlement value and medical incident. This is due, in part, to the confidentiality requirements that are typically attached to a settlement. Such requirements render claim and outcome data less accessible to public health researchers. In the Victorian public health system, all health services are insured by the Victorian Managed Insurance Agency (VMIA), a public insurer wholly owned by the State Government of Victoria.⁵⁶ Therefore, the Victorian public health system is well positioned to access and use insurance data for the purpose of safety and quality improvement. In this regard, the VMIA is similar to the UK’s NHS Litigation Authority (NHSLA). The NHSLA is actively involved in reporting safety learnings to the NHS, publishing reports on, for example, maternity claims,⁵⁷ stillbirth claims⁵⁸ and claims related to medical devices and equipment.⁵⁹ Similarly, the VMIA has published a report highlighting ‘emerging medical indemnity risk’⁶⁰ and provides individual health services with reports on their own drivers of medicolegal claims on request. However, there is no consistent shared reporting. Further, no benchmarking report of litigated outcomes across hospitals is available to hospitals, probably due to the sensitive nature of information. It is possible that at the system level cases would have sufficient similarities to provide direction to policy makers and healthcare managers regarding the key safety shortfalls affecting the health system. In line with Targeting Zero,⁸ which explicitly calls for systematically shared data, the apparent greater appetite for government agencies to better share insights on system-wide emerging risks heralds new opportunities for a more insightful and proactive harm reduction approach than has been not possible until now.^8,61

There may also be opportunities outside of government. Currently, there is neither a requirement for law firms acting for public hospitals or their insurers to provide safety and quality insights nor any routine collation or distribution of their learnings to guide quality improvement. Legal teams acting for medical insurers should be encouraged, perhaps even required as part of their engagement, to routinely comment on potential quality improvements following litigation. They possess intimate knowledge of the health service weaknesses and failures leading to the outcomes they are defending. They also have the benefit of an ‘outsider’s perspective’. Indeed, public inquiries of health services including both Bristol inquiries,^12,62 the Mid Staffordshire inquiry,¹³ the Campbelltown inquiry^63,64 and the Queensland Public Hospitals inquiry⁶ were conducted by lawyers, not doctors. Because they are not dependent on the health system for employment, lawyers perhaps have greater freedom to appraise and evaluate the system more critically without fear of repercussions or professional criticism. Such an approach would require sufficient protections for both doctors providing evidence and the legal teams running the case in order to protect the medicolegal relationship. Nonetheless, not to actively seek feedback from independent legal teams defending public health services, separate to the internal health service corporate counsels, as one method of identifying health service safety and quality gaps seems a safety systems oversight. The benefit of this information would be yet further enhanced if a jurisdictional approach to the collation and dissemination of these learnings were undertaken.

When discussing the judicial system and its applications in guiding healthcare quality and safety reform, it would be remiss not to discuss the role of the Coroners Court. Although not technically a component of the civil judicial system, the Coroners Court is a specialist court tasked with the investigation of reportable deaths. It is an important authority for identifying health service failings.^65,66 The Coroner aims to identify the cause of death, whether death was preventable and what steps can be taken to prevent similar deaths from occurring in the future.⁶⁶ However, these findings are not binding. Although inquest findings are required by the Coroners Act 2008 (Vic) to be made publicly available, there is no formal process for the routine consideration and/or implementation of the Coroner’s recommendations in the quality improvement process. There are certainly opportunities for closer working relationships between the Coroners Court and government safety agencies. However, in the Victorian jurisdiction, informing quality improvement in obstetrics based on the Coroner’s findings faces an additional complication beyond those experienced by other specialties. With specific regard to obstetric deaths, stillbirths fall outside the jurisdiction of the Coroner as specified in the Public Health and Wellbeing Act 2008 (Vic)^65–67. In Victoria it is the responsibility of the Consultative Council on Obstetric and Paediatric Mortality and Morbidity (CCOPMM) to investigate perinatal deaths.⁶⁷ Now administratively supported by Safer Care Victoria, the state’s new quality and safety agency, the CCOPMM maintains a data collection and recording unit that is tasked with the collection of obstetric and perinatal morbidity and mortality data.⁶⁷ This unit is responsible for studying, researching and interpreting all data related to Victorian births. These data are reported to the Department of Health and Human Services, and are used to inform service enhancements in Victoria.⁶⁷ Unlike findings from the Coroners Court, which are publicly released, the investigations and findings of CCOPMM on individual cases are considered in camera and are protected under qualified privilege. As such, findings on individual cases are not discoverable by patients, doctors, health services or regulatory authorities,⁶⁷ unless CCOPMM determines that there is a need for specific referral to the Secretary of the Department of Health and Human Services, as occurred in the case of Djerriwarrh Health Service.⁸ However, CCOPMM findings on individual patient care, even where it was determined that such deficiencies led to avoidable harm, cannot be easily externally shared, either with government departments or the health services or individuals involved, because qualified privilege does not extend to that sharing. Instead, CCOPMM seeks to share its findings through recommendations within its annual reports. Therefore, the routine involvement of CCOPMM in healthcare quality improvement strategies is limited to these summary recommendations rather than specific, targeted recommendations to healthcare providers. In essence, there is an opportunity for summary system learning, a long learning loop, but no opportunity for the short loop of feedback to the individual practitioner. This is in distinct contrast with the Victorian Audit of Surgical Mortality (VASM). VASM was established to allow expert review of all deaths of patients under the care of a surgeon, and the reporting back of findings to the individual surgeon as an educational and quality improvement service. Both the case review and the reporting back of the findings are conducted under the protection of qualified privilege as per the Health Insurance Act 1973 (Aus).⁶⁸ This provision both encourages surgeon participation in the investigation process, in the knowledge that any information disclosed is not accessible otherwise, and allows expert review findings to be shared with the individual surgeon for his or her personal professional quality improvement.⁶⁸ Such an approach of partnering with practitioners to investigate their own failings in care yields obvious benefits for individual practitioners and, in turn, their future patients. However, a significant limitation of this approach to investigation is the restriction on the use and disclosure of these data to people other than those involved directly as the practitioner or peer reviewer of the case. It is, in effect, the opposite of CCOPMM. The individual has opportunities for learning, but the system less so. For system-wide learning to be successful, there is a need to strike a balance between these two differing approaches and effectively deliver both.

Having defined the various actors in the medicolegal process, and opportunities for data sharing in guiding safety and quality reform, we will now consider the specific requirements for these data and comment on how they may be analysed.

Whether insights are to be gained from collated patient complaints and/or medical litigation experience at either a local health service level or at broader state or national system level, a detailed understanding of the complaints or claims will be required to allow sufficiently detailed analysis and presentation. For example, it would be inadequate to simply report that obstetrics is a leading cause of medical litigation.^17,29,69 Such a superficial level of reporting does not inform targeted improvements in care provision. In contrast, the finding that severe perineal trauma is an increasing cause of litigation affords services and clinicians an opportunity to focus on specific improvement initiatives and to measure the effects of the initiatives.⁶⁰ If data are to be shared across health services, and between services and the various agencies (insurers, complaints commissioners, coroners, practitioner boards, departments of health), it will also be necessary that databases are developed or revised to ensure shared taxonomies and data dictionaries. In that regard, the Australian Commission on Safety and Quality in Health Care (ACSQHC) is arguably well placed to lead the development of both a national taxonomy and of data reporting systems to allow sharing and comparison of data across states and territories. The ACSQHC could take responsibility for the on-going custody and analysis of data collected and collated by individual jurisdictions, perhaps publishing an annual atlas, much as it currently does for clinical variation.⁷⁰ Under such a model, we would suggest the initial data collection and collation occurs at a state level using a nationally agreed taxonomy. This could be done by either a lead quality and safety agency, such as Safer Care Victoria or the Clinical Excellence Commission (Fig. 1), or by state health departments, with subsequent central reporting to the ACSQHC for a national report (Fig. 2). This would afford useful insights into national trends and comparisons while still allowing individual jurisdictions an agility to report and respond to local issues. A similar model to this already exists for jurisdictional and national birth outcome reporting.^71–76 The key to ensuring success of a national approach will be to agree a uniform taxonomy from the outset. As a starting point, core datasets could be usefully informed by the research literature that has identified important factors associated with adverse outcomes. Reader et al.²⁵ recently conducted a systematic review to inform their development of a uniform taxonomy for analysing patient complaints with a focus on patient safety, and concluded that a standardised taxonomy for interpreting patient complaints could assist in identifying gaps in patient safety. This taxonomy has been trialled on a subset of 138 complaints to the New South Wales HCC, who concluded that a uniform taxonomy for analysing patient complaints was useful as a method to direct safety and quality improvement.²⁶ With regard to a shared claims taxonomy, patient and practitioner factors, as well as the varying risk exposures of different speciality groups, have all been identified as contributing to the occurrence of medicolegal events.^17,69,77 The legal profession has a long history of using risk factors, such as the type and mechanism of injury, and specific cultural or social traits of the plaintiff to determine the probability of a case succeeding.³¹ This has been necessary for assessing the financial viability of running a case, particularly in a ‘no-win, no-fee’ environment.³¹ Indeed, multiple studies have confirmed the existence of certain ‘high-risk’ medical specialties. These fields attract more litigation and complaints both in number and dollar amount,^{15–17,29,30,78} and primarily consist of procedural craft groups.^17,69 In essence, lawyers calculate the likelihood of successful litigation, derived from an experiential outcome framework, to inform decisions about progressing a case or not. In this regard, healthcare would appear to significantly lag behind the legal profession in the use of case-based factor and experiential learning. We believe that identifying factors that influence a plaintiff’s decision to make a claim shows promise as a method of reducing medicolegal risk exposure.⁷⁹ If these risk factors can be identified and their relative contribution to the occurrence of a medicolegal event can be determined, then it may be possible to predict these events. This would enable targeted interventions to prevent the progression of a claim or complaint, thus lessening the financial burden of litigation and improving quality of care.^31,55 More importantly, an improved understanding of risk factors for medicolegal claims and complaints has applications at a grander scale. In identifying features common to claims and complaints, health systems could act prospectively, taking a continuous improvement approach to healthcare evolution based on learnings from prior adverse events.

Fig. 1.  Flowchart of the proposed complaint-handling framework.

Fig. 2.  Conceptual diagram of the proposed framework.

There are some barriers to building and implementing the systems we propose. First, the various components of both the legal and complaint management systems were purpose built to manage only a single problem. They were not designed to provide a systems overview or to integrate with other reporting mechanisms to derive a shared view of the quality landscape. If complaints and claims experiences are to be shared and analysed to inform future improvements, then definitions, terminology and classifications will require an overhaul and unification. This is where state or federal quality and safety agencies can usefully act to lead and coordinate data dictionary rationalisation, as recommended by Duckett et al.⁸ Second, the ability for some agencies, such as AHPRA and the HCCs, to share data may require some legislative change. However, we suggest that data would be deidentified at source for patient, health practitioner and health service, and still allow useful lessons by extracting only the contributory factors and system classifications. This would hopefully assuage concerns about confidentiality and reputational risk. Third, building the capability will, of course, require investment. We would argue that if shared insights from claims and complaints can contribute to improved quality and reduced harm, the return on investment will be considerable. In this regard, the effect of a prospective ‘experience mapping’ approach would likely be greater than acting on potential claims and complaints at the individual patient or health service level, both for risk management and quality of care. As the adage goes, ‘prevention is better than cure’. Because medicolegal claims and complaints represent the most serious breaches in patient safety, it is surely the responsibility of health system regulators to take every opportunity to prevent them.

Australian state health systems already have the key components and structures that would allow shared learnings from patient complaints and medicolegal claims to be used more effectively and purposefully to inform quality and safety improvement. However, the various components, whether patient complaints handling processes, central complaints mediation services, hospital legal services or the public insurer, were neither principally established nor are currently linked in a manner to systematically service an improvement agenda. We believe that the continued occurrence of whole-of-service failures will not likely be prevented in the future by a single service approach. Rather, a whole-of-system solution affording routine data sharing and analysis, as recommended in Targeting Zero,⁸ and reporting of data across and between agencies is required. We need to learn from each other’s mistakes, benchmark ourselves against each other and do both in a timely and informed manner. Such an approach to shared learning in combination with incident reporting and open disclosure then needs to be followed-up with system-wide and measureable action. We need to share the pain. Only then can we hope to prevent it in the future.

The authors declare no conflicts of interest.