OUP user menu

Look back and talk openly: responding to and communicating about the risk of large-scale error in pathology diagnoses

Rosemary Aldrich, Peter Finlayson, Kim Hill, Margaret Sullivan
DOI: http://dx.doi.org/10.1093/intqhc/mzr084 135-143 First published online: 11 January 2012


Quality problem or issue In 2005 we were informed about the possibility of a systemic error in pathology diagnoses of ∼7400 histological tests done between 4 and 6 years earlier.

Initial assessment We determined to undertake a lookback and apply principles of open disclosure to inform the affected community of ∼200 000 people.

Choice of solution The lookback included subjecting all cases to independent pathology review. The public announcement of the review included an unreserved apology and took place before the results of the re-examination of the pathology specimens were known.

Implementation The lookback involved the simultaneous implementation of five critical elements: leadership and governance, risk assessment and planning, implementation of the independent review, procedures for patient care and communication and open disclosure. Protocols were developed to care for those patients whose original test results were found to be incorrect.

Evaluation The original result for >200 patients was incorrect, and 38 had experienced clinical consequences. There was no public panic as a result of the wide open disclosure. Few related legal claims or complaints were made. The impact of the pathology diagnostic error has continued to 2011 for some patients.

Lessons learned Openly disclosing a risk of widespread error meant the community could be supported with information and medical management as needed. Credentialing and peer-review processes for senior staff must be precise and collegiate. Sometimes action has to take place even when the risk is ill defined. There are five critical elements in planning and implementing a large-scale lookback.

  • lookback
  • pathology
  • error
  • open disclosure
  • apology
  • patients and community


In March 2006 Hunter New England Health (HNE Health), a public health organization in New South Wales (NSW) Australia, announced that ∼7400 histological and cytological cases reported 4–6 years earlier were to be re-examined by independent pathologist reviewers. The independent review was implemented in response to information that the pathologist who reported the cases had returned a ‘much higher than expected error rate’ in his work in another public health organization. The results of the independent review were made public 5 months later. In this paper we aim to both report the results of the independent pathology review and describe the planning and implementation of the elements which together comprised the wider ‘lookback’ (of which the independent pathology review was one element). Lookback is defined for our purposes as a process triggered by an incident or concern in the health care of a group of commonly affected patients [1]; the term has been used to describe tracing recipients of potentially infected blood [2].

Australia has had a universal public health-care system since 1984. HNE Health provides comprehensive acute, community, mental and population health services (including pathology and imaging) for the 850 000 people who live in a region bigger than England, from the Pacific coast to rural and remote parts of north-western NSW. It was established in January 2005 through an amalgamation of three smaller NSW public health organizations.

In July 2005 the HNE Health chief executive was informed by the most recent employer of an anatomical pathologist that an audit had been undertaken of the pathologist's work after a routine peer review of 20 case reports had prompted further examination. More than 6000 cases reports were examined by two or more anatomical pathologists. The audit had revealed a 3.3% ‘significant variation rate’, meaning that 1 in 30 patients was found to have a diagnosis which so varied from the original that the independent pathologist reviewers considered that the original diagnosis may have resulted in clinical sequelae or sub-optimal care. HNE Health was told of the audit because the pathologist had been employed by one of the HNE Health's predecessor organizations between 1999 and 2001. Notwithstanding the lack of clarity around ‘expected’ error rates in anatomical pathology [35], we were compelled to subject all cases reported by the pathologist to re-examination by independent experts. We began planning the independent review.

We found few reports on planning an extensive independent review triggered by a risk of error. We drew on our collective experience in public health and health services management, and consulted others who had undertaken similar tasks. Internationally, the New Zealand experience was helpful. In response to a 1999 High Court finding that a pathologist had misdiagnosed a patient's cervical smear, resulting in that patient progressing to cervical cancer, the New Zealand Health Funding Authority (HFA) initiated an unblinded review of all smears reported by the pathologist between 1991 and 1996, comprising almost 23 000 reports affecting >12 000 women [6]. The primary purpose of the HFA re-reading was to ‘identify and address any risks to the health and well-being of women whose slides had originally been read and reported’ by the pathologist in question. The aim of our lookback was to quantify, communicate ethically with our community about, and care for people affected by the risk of wide-spread error in cytology and histopathology reporting.


The conduct of the HNE Health pathology lookback required simultaneous attention over time to five interacting elements, as set out in Table 1: leadership and governance, risk assessment and planning, implementation of the independent review, procedures for patient care and communication and open disclosure.

View this table:
Table 1

Timeline for planning and implementation of the lookback

Leadership and governanceRisk assessment and planningImplementation of the independent reviewProcedures for patient careCommunication and open disclosure
July 2005Initial notification
August 2005–January 2006Decision to undertake lookbackNature of risk identifiedProtocol for independent review developed, peer reviewedProcedures for patient care developed in consultation with senior cliniciansPlanning communication campaign
Lookback team establishedLiterature review to determine best practice
Pathologist's qualifications verifiedCases reported by pathologist identifiedContracts with independent pathology reviewers negotiatedDrafting materials and fact sheets
Communication with state public health and registration authoritiesReports and specimens for each case collatedDaily communication between lookback team membersBriefing reports for authorities
Data base developed
Each case given a unique identifier
February 2006High-level planning for public announcementCollated tests and results transported to independent reviewersEstablish toll-free number
March 2006Chief executive makes public announcement of independent pathology review and addresses questions from mediaChief executive briefed weeklyIndependent reviewers commence reviewComplaints and concerns addressedFirst public open disclosure
Media releases
Media interviews
Toll free number live
Progress and risk assessment reviewed monthly by lookback team with chief executive
Implementation and evaluation issues managed as they arise
April 2006First results receivedSignificant variation protocols enacted
Affected patients traced and contacted
Affected patients and carers assessed by their doctors, offered clinical and psychological care as needed
August 2006(Late August) Chief executive makes public announcement of review resultsCurrent processes for peer-review and collegiate support for pathologists reviewed(Early August) Last results received as per contractual arrangementsCare for affected patients and carers continues as determined by their cliniciansFinal public announcement
September 2006–March 2007Conduct of lookback reviewedContractual obligations finalizedComplaints and concerns addressed
Lookback protocols revised in light of experienceFollow-up for affected patients and carers continues as requiredToll-free number closes March 2007
To presentNSW Health Lookback policy and HNE Health lookback policy compliance procedure published September 2007Follow-up for affected patients and carers continues as required

Leadership and governance

A lookback team was established, comprising a project leader (public health physician), project officer (senior nurse manager) and the medical director of the facility at which the pathologist had worked. Clear reporting lines were established. The project leader reported the lookback's progress to the chief executive weekly, and raised issues with the Director of Clinical Governance as required. In addition, the organization's Director of Communications was closely involved, planning and managing public communication aspects of the lookback. Legal advice was sought only in relation to contracting independent pathology reviewers. Appropriate state public health and regulatory authorities were informed of the matter and the intention to undertake the lookback.

It was established through direct contact with the American Board of Pathologists that the pathologist had gained his Board qualifications in anatomical pathology several years earlier; this qualification permitted him to charge the Australian government for services provided to patients, although did not qualify him as a specialist pathologist with the Royal College of Pathologists Australasia (RCPA). As the pathologist's whereabouts were unknown, the NSW Medical Board was asked to inform the pathologist that a lookback was commencing.

Risk assessment and planning for the independent review

The number of cases and the nature of the risk informed the scope of the independent review. The laboratory provided the only public pathology services for a population of ∼200 000 people resident in numerous rural cities, towns and communities. The pathologist had reported 7432 cases (7329 histological and 103 cytological), at least 90% of all pathology reported through the laboratory over his 2-year appointment. The pathologist diagnosed or excluded pathology of skin, breast, prostate, upper and lower gastrointestinal tracts and reproductive and other organs, but had not reported urine, blood, sputum or marrow specimens.

Independent reviewers from three separate large pathology laboratories were contracted and a time frame for the return of results set. The independent review protocol was developed in consultation with senior laboratory scientists and peer reviewed by the RCPA as adequate for the purposes of a patient-focused review. The protocol required that each case would be examined by an independent reviewer, and noted as either agreed, or as a minor variation (a variation in some aspect of the diagnosis but with no potential for impact on patient care) or a significant variation (where peer-reviewed diagnoses varied from the original to the extent that it was considered the original diagnosis might have led to avoidable clinical sequelae or suboptimal care). Every significant variation was then reviewed by between two and seven independent pathologists and an amended report issued.

Implementation of the independent review

More than 55 000 specimen slides which comprised the 7432 cases to be reviewed were identified, verified as labelled correctly, packaged securely, collated with printouts of the original pathology reports and delivered by a staff member to the laboratories of independent reviewers. Reports of significant variations once finalized were sent by the independent reviewers to the project team electronically. At the conclusion of the review all slides were returned to our laboratory without breakage, accompanied by a hard copy summary notation (agree, minor variation or significant variation) for each case, signed by the reviewer. The results for each case were entered into the patient's records, either when received (for significant variations) or at the conclusion of the independent review.

Procedures for patient care

Informing and providing care to patients whose review diagnosis varied significantly from their original diagnosis was undertaken where possible by the clinician who had requested the original test. Most clinicians still worked in the area and were very willing to be involved in follow-up care when approached. The independent reviewers undertook to inform the lookback team immediately a significant variation was identified, at which time the clinician who had requested the original test was contacted and agreement reached about whether the clinician or the lookback team would contact the patient in the first instance. By managing communication with affected patients in this way we could ensure that all patients with significant variations to their original diagnosis would be contacted.

Patients who had moved from the address recorded at the time of their original pathology test were traced through telephone directories and electoral rolls. Where the patient had moved away from the local area or where the clinician who had requested the original test had left practice the patient was contacted and asked to nominate a general practitioner. The lookback team then contacted and asked the nominated general practitioner to make an assessment and provide care. Where the patient had no general practitioner, assessment and clinical care was arranged with an appropriate medical practitioner at a location convenient to them.

Doctors of patients with significant variations assessed whether their patient had experienced clinical sequelae or sub-optimal care as a result of their original diagnosis. HNE Health undertook to fund any treatment needed as a consequence of the original diagnosis. Any conversations between members of the lookback team and patients with significant variations were confirmed in writing. The emphasis on patient care and the provision of patient-focused information remained throughout the conduct of the review.

Communication and open disclosure

We determined to communicate openly with patients and the community. A public announcement about the independent review was made after testing began but before any results were known. On the day of the announcement, the chief executive first informed community leaders and local politicians, and then met with senior health managers, clinicians and other health service staff. A press conference took place immediately afterwards to ensure that maximum information was made available to the community that same day; a toll-free phone number was publicized on that evening's local television news. Anticipating some public alarm additional mental health personnel were on standby from the day of the announcement. The principles of open disclosure [7] informed the announcement: a description of what had occurred, an apology for distress the announcement might cause, details of the investigation underway and what would happen in the event of a significant variation being identified (and that this would be at no cost to the patient). The toll-free telephone number was staffed by clinical members of the lookback team from 8 a.m. to 10 p.m. daily for 1 year. Anyone who wished to contact the lookback team was encouraged to use the toll-free number or use an email address also publicized.

Once cases had been reviewed, and patients with significant variations informed, in August 2006 the results of the independent review were announced, a further apology given and inquiries directed to the toll-free number. All patients (or their nominated carer where the patient had died) included in the review were sent a concluding letter containing a written apology and confirming their result as either a significant variation, or not. Both the toll-free number and a health service number were included in the concluding letter so that patients could know how to obtain further information in the future.


Results of the independent review

Figure 1 shows the results of the review. In 2.9% (n= 217) of cases the diagnosis of at least two independent reviewers ‘varied significantly’ from the original diagnosis, that is, the independent pathology reviewers considered that in 217 patients the initial (incorrect) diagnosis had led potentially to incorrect clinical management; Table 2 shows the distribution across organ systems of these significant variations. We traced 214 of the 217 patients with significant variations; of the 214 their doctors told us that in their opinion 38 patients had experienced clinical sequelae as a result of their original diagnosis, ranging from having had unnecessary procedures (over-diagnosis) to having experienced delayed diagnoses or treatment of various conditions (underdiagnosis). Of the 214 patients traced, 45 had died since their original test. It was not possible to determine definitively whether anyone had died as a result of their misdiagnosis.

View this table:
Table 2

Significant variations related to specific organ

OrganNo. of significant variations% of total significant variations
Gastrointestinal pathology, upper3114.3
Gastrointestinal pathology, lower2612.0
Gynecological pathology2612.0
Breast pathology188.3
Urological pathology115.0
Respiratory pathology83.7
Endocrine pathology62.8
Body cavity pathology20.9
ENT pathology20.9
Renal pathology10.5
Figure 1

Results of independent review of 7432 anatomical pathology diagnoses, Hunter New England Health, NSW Australia 2006. Note: Minor variation where independent reviewers disagreed with the original diagnosis but believed that there was no risk of clinical sequelae or suboptimal care resulting from the original diagnosis. Significant variation where independent reviewers disagreed with the original diagnosis and considered that there was a risk of clinical sequelae or suboptimal care resulting from the original diagnosis.

Community response to the initial open disclosure communication strategy

There was a rapid response to the public announcement in March 2006 that an independent review of a large number of pathology cases was to be undertaken. The first phone call to the toll-free telephone number was received moments after the number was broadcast on the evening news. Local newspaper reportage and talkback radio the next day gave valuable feedback from non-organizational sources about the approach taken, including the view expressed by a health professional unrelated to our organization that HNE Health had been ‘responsive and responsible’ in undertaking the review. Mental health managers briefed beforehand reported that there was no increased demand on community mental health services in the days following the announcement; additional mental health professionals who had travelled to the city to support local services were returned to their usual positions after 2 weeks.

Community use of the toll-free telephone and email

There were over 500 calls to the toll-free telephone number between March and October 2006, of which >300 were initiated by members of the public and local health professionals, as set out in Table 3. No results were given over the phone; results were returned to patients or families via their own doctor or an appointment with a lookback team doctor. The dedicated email account received one email only. There were a further 47 calls to the toll-free number until it was closed in March 2007. HNE Health has received a further 34 phone inquiries since March 2007 in relation to the review.

View this table:
Table 3

Timing and subject matter of telephone calls to the toll-free telephone number, March–October 2006 inclusive

MonthEventNo of callsSubject matter and responses
MarchPublic announcement and open disclosure126Callers wishing to determine whether/make sure they were included
Callers seeking more information about type of tests
Some callers very worried, requesting to have their cases reviewed as a matter of urgency (arranged to allay their concerns; treating doctors also involved in managing concerns and distress; callers offered counseling—one accepted, one declined
Includes three doctors seeking to identify which of their patients included in review
AprilFirst significant variations reported8Callers seeking more information, inquiring about their results, wishing to speak to additional doctors
Public media reported ‘results were back’Callers seeking assistance with understanding their follow-up arrangement; assured they would be contacted if their results were of concern
MayResults returned to patients7Complaint made initially via toll-free number
JuneResults returned to patients6Callers seeking more information
JulyResults returned to patients22Callers seeking more information
AugustFinal results reported publicly at end August18Callers seeking confirmation that they did not have a significant variation
SeptemberPublic awareness of review and results87Callers seeking more information
OctoberContinuing support for community38Callers seeking more information

Community response to the results of the independent review

Reactions from patients and other members of the community were varied. Two people were extremely distressed by the review results (one whose partner had a significant variation and another whose relative had a minor variation), and both were given on-going mental health support. There was an increase in calls to the toll-free telephone number in the weeks following the public announcement of the results (Table 3). There were two legal actions (both resolved) and two complaints made about the matter. One of the complaints was identified when a patient's experience was the subject of a media report. However, that patient told us when contacted that an acquaintance had solicited media attention for his story without his permission. Each concern raised with HNE Health was discussed by bringing together the patient and/or their family with a senior doctor and manager. HNE Health undertook to pay for any medical care required as a consequence of a person being included in the independent review, including for mental health issues. To our knowledge treatment for pathology review-related matters and access to mental health support continues for two patients and their families.


We used explicit principles of open disclosure about health-care error to inform the timing and nature of public communications about the lookback described in this paper. There is wide agreement that central to safety and quality in health is the need for openness about error [710]. However, principles of openness about error in health care have not in general considered how to communicate with a wider community when a risk of potentially far-reaching error is identified [11].

Apologies given about failures of governance in health care have been given typically after a matter becomes public. In 1992 (some 5 years after the conclusion of its Ministerial Inquiry), New Zealand hospital authorities apologized to women affected by a 20-year cervical cancer ‘experiment’ at the National Women's Hospital Auckland, New Zealand, in which some women diagnosed with cervical abnormalities were left untreated so that ‘researchers’ could observe the natural history of the condition [12, 13]. The Premier of Newfoundland and Labrador apologized in 2008 for Eastern Health's failure to disclose its 2005 knowledge of erroneous breast hormone receptor tests results for >1000 women [14]. Our 2006 strategy of proactive disclosure, an apology and a ‘call-in’ centre staffed by experts aligns closely with recent recommendations about the conduct of large-scale disclosure [15]. As well as being the ethical thing to do, providing timely accurate information about risk or hazard has been shown to reduce likelihood of public catastrophizing and outrage [8, 16]. While recognizing that publicly discussing the risk of widespread error will necessarily take into account the specific national and health policy context, our experience serves as a key lesson to illustrate that it is possible to openly disclose matters of potentially widespread health-care error without causing widespread panic.

A second lesson from this work was that credentialing, supervision and peer-review processes for senior staff must be precise and collegiate. Our investigation into the circumstances which had years later given rise to the lookback found that on appointment the pathologist had been credentialed to work under the supervision of an external pathologist until he was able to practise independently, the assumption being he could practise independently (including billing for services) only once he obtained his Australian specialist pathologist qualification. However, unknown to the credentialing panel, the pathologist's American certification qualified him to charge patients for his services, and once this had occurred, the external billing facility provided by the supervising pathologist was not required and supervision ended, leaving the pathologist in solo practice. With respect to peer review, the pathologist had participated in external quality assurance exercises confidentially, meaning that results of testing were never made known to his employer or peers. At the time of the lookback it was established that robust processes for peer support, quality assurance and peer review had been implemented in the intervening years.

A third lesson from our lookback concerned how to deal with ill-defined risk. Literature suggested that pathologists made misdiagnoses with a significant clinical impact warranting an amended report <1.5% of the time [3, 4]. A review of a UK locum histopathologist found that in <0.1% of his diagnoses—7 out of >7000-patients ‘had suffered serious consequences as a result of an incorrect reading of their histology samples’[17]. The UK Royal College of Pathologists reviewed 90 000 slides in a cervical smear audit in 1999 and found a cancer misdiagnosis rate of 1.6% [18]. There were no data on the error rate of pathologists in Australia. In our case at the laboratory of his most recent employment the pathologist had undergone the same peer review as his colleagues, and his results alone had prompted a larger audit. Once informed about the pathologist's established significant variation rate we were obliged to act, although this incurred substantial cost as all tests previously reported had to be re-examined, and staff resources would be devoted to supporting the lookback.

The fourth important lesson was of the experience of undertaking the lookback itself, which took about a year to plan and implement from first alert to final announcement. Having identified and planned and implemented the critical elements of a large-scale lookback will assist us should there be a need to undertake or advise in a similar matter again.

Our experiences informed subsequent development of the NSW Health Lookback policy [1], and have motivated us to report our processes and findings to assist others faced with the planning, communication and logistical challenges of a large-scale lookback. Importantly, we demonstrated that talking openly about the lookback meant that patients and their families could be supported as results indicated they should be. We suggest that our experience with applying principles of open disclosure demonstrates the utility of the model in meeting our ethical obligations to the communities we serve, as well-being the means of providing the best clinical and supportive outcome for patients and carers affected by adverse events.


The investigation described in this report was funded by Hunter New England Health.


The authors acknowledge the contribution to the Lookback of former HNE Health Chief Executive, Mr Terry Clout, former HNE Health Communications Director Ms Allison Bone and inaugural Director HNE Health Clinical Governance Unit Professor Allan Spigelman; former HNE Health pathologist Dr Lance Meng; and staff of South East Sydney Illawarra Pathology for their contribution to and advice in the conduct of the Lookback. In addition, the authors are indebted to the many clinicians of the New England region of NSW without whom the patients and families affected by the Review could not have received the care they did.


  • This paper is based on a brief presentation given at the International Society for Quality in Health-care meeting, Boston, October 2007.


View Abstract