Identifying Patients at Risk of Not Receiving Timely Community Mental Health Follow-Up After Psychiatric Hospitalisation Using Linked Routinely Collected Data.

Timely receipt of community-based follow-up after inpatient psychiatric discharge is associated with positive outcomes. This retrospective cross-sectional study aimed to identify socio-demographic and clinical factors associated with failure to receive community mental health follow-up within 7 days. Routinely collected hospital and community mental health data were linked for all inpatients discharged with a mental health condition in 2017 to 2019 in a local health district in New South Wales, Australia. Of the 8780 patients discharged, 28% (n = 2466) did not have 7-day follow-up. Males were significantly more likely than females to fail follow-up. Adjusted logistic regression analyses revealed that both male and female patients aged 65 years and older were generally less likely to fail follow-up than those who were younger; conversely, patients referred to a hospital by a law enforcement agency and those who left the hospital at their own risk were more likely to fail follow-up. Other factors significantly related to failure to follow-up varied between the sexes. Improved outcomes may be achieved by enhancing the transition from inpatient to outpatient care through targeted strategies aimed at patients who are more likely to disengage with care.

Factors associated with mental health representations to the emergency department within six months.

BACKGROUND/OBJECTIVE: ED representation places a tremendous drain on resources with mental health (MH) representation among the most common. This study aimed to identify patient and clinical factors associated with 28-day and six-month ED MH representation of an index MH ED presentation. METHOD: All MH related ED presentations from 1 January 2017 to 30 June 2019 were extracted from routinely collected administrative data. Cox regression and multinomial logistic regression models tested associations between patient characteristics and risk of representation. RESULTS: For the 8,010 patients, 28-day and six-month representations were 8 % and 16 % respectively. Self-identifying with a MH problem at index presentation (28-day hazard ratio (HR) = 1.48, 95 % CI = 1.19-1.84; six-month HR = 1.52, 95 % CI = 1.29-1.78), leaving ED before completing treatment (28-day HR = 4.13, 95 % CI = 3.36-5.08; six-month HR = 2.52, 95 % CI = 2.12-2.99), no private health insurance (six-month HR = 1.34, 95 % CI = 1.08-1.66), and hospital admission within one year prior to index (six month MH-related admission vs non-MH, HR = 1.59, 95 % CI = 1.19-2.13) were associated with increased risk of representation. Being uninsured was associated with frequent six-month representation among adults aged 16-39 years (OR = 3.16, 95 %CI = 1.59-6.25). CONCLUSION: Self-identifying with a MH problem, leaving ED prematurely, being uninsured and prior hospitalisation are areas for in-depth investigation for improved understanding of unplanned representations.

Sociodemographic characteristics and clinical outcomes for people presenting to emergency departments with mental health diagnoses.

BACKGROUND: Hospital emergency departments (EDs) are experiencing a growth in presentations with mental health (MH) diagnoses. AIM: Describe and compare sociodemographic characteristics and clinical outcomes for people with MH and non-MH diagnoses. METHODS: A retrospective study examined routinely collected data for ED presentations in a health district in western Sydney, Australia from 2016 to 2019. Regression models examined variables according to MH status, overall and by age. RESULTS: Individuals with MH diagnoses accounted for 3.4% of 647,787 ED presentations. MH presentations were most commonly female (51.5%), aged 16-39 years (62.5%), arrived after hours (60.3%) and via ambulance (52.8%). MH presentations were more likely to be triaged category 2 (OR = 1.58,95%CI = 1.54-1.63) and not seen on time (OR = 1.20,95%CI = 1.17-1.24). They had higher odds of a longer ED stay (OR = 1.96,95%CI = 1.90-20.1), after which they were less likely to be admitted (OR = 0.56, 95%CI = 0.55-0.58) and more likely to be transferred (OR = 3.81,95%CI = 3.66-3.97) or leave before treatment was completed (OR = 1.83,95%CI = 1.74-1.92). CONCLUSION: Characteristics and outcomes for people presenting to ED with a MH diagnosis significantly differ from those without a MH diagnosis. Provision of timely care is a particular concern. Identifying causes for delays within and external to the ED, and implementing targeted strategies to ameliorate them are required to optimise care.

Caesarean sections before and during the COVID-19 pandemic in western Sydney, Australia.

BACKGROUND: To determine the changes in emergency and elective caesarean section (CS) rates since the COVID-19 pandemic, identify the groups most affected, and examine changes in the factors associated with CS rates, and reasons for CS. METHODS: We conducted a retrospective cohort study using routinely collected data of 22,346 births from before the pandemic (January 2018-February 2020) and 18,597 births during the pandemic (March 2020-December 2021). Data were analysed using multinominal logistic regression. RESULTS: The CS rate increased by 4.1% (from 30.1% to 34.2%), reflecting increases of 2.3% in emergency CS (from 11.5% to 13.8%) and 1.7% in elective CS (from 18.7% to 20.4%). Large groups with notable increases were women who were nulliparous (7.2% increase), from South Asia (6.0%), obese (5.2%) and giving birth at a small hospital (6.1%). Compared to pre-pandemic, the relative risk of an emergency CS versus a vaginal delivery increased 1.36 times (adjusted relative risk ratio (aRRR) = 1.36; 95% CI = 1.27, 1.45) and the risk of having an elective CS increased 1.11 times (aRRR = 1.11; 95% CI = 1.04, 1.20). Factors associated with both emergency and elective CS were age, region of birth, reproductive history, body mass index, hypertension, diabetes, mode of antenatal care and hospital. Socio-Economic Indexes for Areas and antenatal care were only associated with elective CS. Baby gender was only associated with emergency CS. Preterm gestation at delivery was associated with reduced emergency but increased elective CS. Foetal compromise was the most common indication for emergency CS (43.2%) and increased the most (8.0%). Previous CS was the most common indication for elective CS (61.5%) and reduced the most (1.9%). CONCLUSIONS: Both emergency and elective CS rates increased significantly during the pandemic, with the former increasing at a higher rate. The persistent upward trend of CS rates, exacerbated by increasing proportions of nulliparous women undergoing CSs, is concerning.

Australia has a very high caesarean section (CS) rate that varies greatly between groups of women with different socio-economic characteristics and reproductive histories. Information regarding changes in CS rate since the COVID-19 pandemic in Australia is limited. We conducted a study comparing CS rate before and during the pandemic, using routinely collected data. Both emergency and elective CS rates increased significantly during the pandemic with emergency CSs increased at a higher rate than elective CSs. Several groups of women experienced large increases in CS rate. Factors associated with and reasons for emergency CSs were different from those for elective CSs. Health services should be prepared to minimise effects of future pandemics on CS rate. To be most effective, interventions to reduce non-medically justified CSs should focus on women who are from South Asia, obese, admitted to a small hospital, and are nulliparous. Different approaches are needed to reduce emergency and elective CSs.

A single-item measure of self-rated mental health and psychological distress. In what situations can a single-item measure be useful?

OBJECTIVE: To examine the association between self-rated mental health (SRMH) and psychological distress (PD) at multiple periods and subsequently assess the potential of SRMH as a screening tool. METHOD: Staff working at a designated COVID-19 hospital in Sydney, Australia during March to May 2020 completed the SRMH and Kessler Psychological Distress Scale (K10) within a larger survey examining the pandemic's impact on health and well-being. SRMH was assessed before the pandemic (baseline), during its first peak in 2020 (time 2) and several months later (time 3). K10 was assessed for time 2 and time 3. All assessments took place at time 3. RESULTS: At time 2, 80% of respondents with high PD and 25% with low PD reported poor SRMH (χ2 = 21.3, p < .0001). At time 3, 90% with low PD reported good SRMH. Risk of high PD was greater for respondents with persistently poor SRMH (time 2: OR = 18.2, 95% CI = 7.7-42.8; time 3: OR = 14.4, 95% CI = 6.9-29.9) and, to a lesser extent, for those whose SRMH declined (time 2: OR = 11.6, 95% CI = 6.6-20.4; time 3: OR = 13.8, 95% CI = 2.9-66.9), compared to those with persistently good SRMH. CONCLUSIONS: During a crisis SRMH can identify the majority of those most likely to benefit from additional assessment and support. Persistently poor SRMH indicates highest risk, detectable from routine screening.

Verification of administrative data to measure palliative care at terminal hospital stays.

BACKGROUND: Administrative data and clinician documentation have not been directly compared for reporting palliative care, despite concerns about under-reporting. OBJECTIVE: The aim of this study was to verify the use of routinely collected administrative data for reporting in-hospital palliation and to examine factors associated with coded palliative care in hospital administrative data. METHOD: Hospital administrative data and inpatient palliative care activity documented in medical records were compared for patients dying in hospital between 1 July 2017 and 31 December 2017. Coding of palliative care in administrative data is based on hospital care type coded as "palliative care" and/or assignment of the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision, Australian Modification (ICD-10-AM) palliative care diagnosis code Z51.5. Medical records were searched for specified keywords, which, read in context, indicated a palliative approach to care. The list of keywords (palliative, end of life, comfort care, cease observations, crisis medications, comfort medications, syringe driver, pain or symptom management, no cardiopulmonary resuscitation, advance medical plan/resuscitation plan, deteriorating, agitation, restless and delirium) was developed in consultation with seven local clinicians specialising in palliative care or geriatric medicine. RESULTS: Of the 576 patients who died in hospital, 246 were coded as having received palliative care, either solely by the ICD-10-AM diagnosis code Z51.5 (42%) or in combination with a "palliative care" care type (58%). Just over one-third of dying patients had a palliative care specialist involved in their hospital care. Involvement of a palliative care specialist and a cancer diagnosis substantially increased the odds of a Z51.5 code (odds ratio = 11 and 4, respectively). The majority of patients with a "syringe driver" or identified as being at the "end of life" were assigned a Z51.5 code (73.5% and 70.5%, respectively), compared to 53.8% and 54.7%, respectively, for "palliative" or "comfort care." For each keyword indicating a palliative approach to care, the Z51.5 code was more likely to be assigned if the patient had specialist palliative care input or if they had cancer. CONCLUSION: Our results suggest administrative data under-represented in-hospital palliative care, at least partly due to medical record documentation that failed to meet ICD-10-AM coding criteria. Collaboration between clinicians and coders can enhance the quality of records and, consequently, administrative data.

Managing COVID-19 in an Australian designated isolation facility: Implications for current and future healthcare crises based on in-depth interviews.

Health care workers' (HCWs) lived experiences and perceptions of the pandemic can prove to be a valuable resource in the face of a seemingly persistent Novel coronavirus disease 2019 (COVID-19)-to inform ongoing efforts, as well as identify components essential to a crisis preparedness plan and the issues pertinent to supporting relevant, immediate change. We employed a phenomenological approach and, using purposive sampling, conducted 39 semi-structured interviews with senior healthcare professionals who were employed at a designated COVID-19 facility in New South Wales (NSW), Australia during the height of the pandemic in 2020. Participants comprised administrators, heads of department and senior clinicians. We obtained these HCWs' (i) perspectives of their lived experience on what was done well and what could have been done differently and (ii) recommendations on actions for current and future crisis response. Four themes emerged: minimise the spread of disease at all times; maintain a sense of collegiality and informed decision-making; plan for future crises; and promote corporate and clinical agility. These themes encapsulated respondents' insights that should inform our capacity to meet current needs, direct meaningful and in situ change, and prepare us for future crises. Respondents' observations and recommendations are informative for decision-makers tasked with mobilising an efficacious approach to the next health crisis and, in the interim, would aid the governance of a more robust workforce to effect high quality patient care in a safe environment.

Pandemic Loneliness in Healthcare Workers. Does It Predict Later Psychological Distress?

Increased loneliness associated with the COVID-19 pandemic has been widely reported, with healthcare workers at increased risk compared to the general population. Pre-pandemic research indicates that loneliness has long-term detrimental effects on mental well-being, but the effect of loneliness in the context of COVID-19 is not clear. We conducted an online survey of healthcare workers (HCWs) at a large teaching hospital in Sydney, Australia after the peak of the first wave of the pandemic in 2020. Over one-third experienced loneliness at the peak of the first wave. An observed association with high psychological distress in subsequent months was attenuated after adjusting for status of mental well-being during the peak and before the pandemic. Mental well-being during the peak of the pandemic and, to a lesser extent, mental well-being before the pandemic were the strongest predictors of later distress. Increased conflict at home was also predictive of later distress. The mental health of HCWs is an important consideration at any time and is especially so in the face of crises such as the current global pandemic. Strategies to enhance baseline mental well-being and bolster well-being during crisis situations should assist HCWs cope with unexpected stressors and reduce the associated detrimental psychological consequences.

Are healthcare workers particularly vulnerable to loneliness? The role of social relationships and mental well-being during the COVID-19 pandemic.

Loneliness is a major public health issue with renewed prominence due to the COVID-19 pandemic and associated social restrictions. Healthcare workers (HCWs) may be at heightened risk, but research is lacking. We measured the prevalence of loneliness among HCWs during the pandemic in 2020 and examined pre-pandemic predictors and pandemic experiences associated with loneliness. HCWs at a designated COVID-19 hospital in Sydney, Australia completed an online survey examining health and well-being before and during the pandemic and changes to work, family and social experiences. Loneliness had negatively affected the well-being of 129 (39%) respondents (n â€‹= â€‹330). Pre-pandemic factors predicting loneliness were younger age (<30years compared to ≥50years), having ever been told you had a mental health problem and living alone. These became non-significant when pandemic-related factors were added to the regression. Less contact with family and friends, increased conflict at home, and living alone or with family but not a partner, increased the odds of loneliness, while a sense of camaraderie with colleagues had the opposite effect. Psychological distress and poor mental health during the pandemic were also positively associated with loneliness. Efforts to promote congenial social contacts may be effective in averting loneliness among HCWs.

Detrimental changes to the health and well-being of healthcare workers in an Australian COVID-19 hospital.

BACKGROUND: Most studies examining the psychological impact of COVID-19 on healthcare workers (HCWs) have assessed well-being during the initial stages or the peak of the first wave of the pandemic. We aimed to measure the impact of COVID-19 and potential changes over time in its impact, on the health and well-being of HCWs in an Australian COVID-19 hospital. METHODS: An online questionnaire assessed current and retrospective physical and mental health; psychological distress (Kessler Psychological Distress Scale); lifestyle behaviours; and demographics, providing measures of health and wellbeing at three phases of the pandemic. Targeted staff were invited to participate via email and in-person. Additional promotional activities were directed to all staff. Changes in general health, mental health and psychological distress were examined using McNemar's Chi-square. Associations between other categorical variables were tested using Chi-Square or non-parametric equivalents as appropriate. Logistic regression explored risk factors for current distress. RESULTS: Four hundred thirty-three eligible HCWs answered all (74 %) or part of the questionnaire. Current self-rated health and mental health were significantly better than during the height of the pandemic, but had not returned to pre-pandemic levels. Psychological distress was significantly more common during the height of the pandemic (34.2 %) than currently (22.4 %), and during the height of the pandemic distress was significantly more common among younger than older HCWs. Females were significantly more likely to be distressed that males currently, but not during the height of the pandemic. High distress during the height of the pandemic was more likely to be maintained by HCWs who were less physically active than usual during the height of the pandemic (OR = 5.5); had low self-rated mental health before the pandemic (OR = 4.8); and who had 10 or more years of professional experience (OR = 3.9). CONCLUSIONS: The adverse effects of the pandemic on HCWs have lessened with the easing of pandemic demands, but health and well-being have not reverted to pre-pandemic levels. This indicates continued exposure to elevated levels of stress and/or a sustained effect of earlier exposure. Initiatives that provide ongoing support beyond the pandemic are needed to ensure that HCWs remain physically and mentally healthy and are able to continue their invaluable work.

Specialist Palliative Care Activity at an Acute Care Tertiary Hospital and Its Representation in Administrative Data.

OBJECTIVE: To quantify and examine specialist palliative care (SPC) in-hospital activity and compare it to routinely collected administrative data on palliative care (PC). METHODS: All patients discharged from a large acute care tertiary hospital in New South Wales, Australia, between July 1 and December 31, 2017, were identified from the hospital's data warehouse. Administrative data were supplemented with information from the electronic medical record for hospital stays which were assigned the PC additional diagnosis code (Z51.5); had a "palliative care" care type; or included SPC consultation. RESULTS: Of 34 653 hospital stays, 524 were coded as receiving PC-based on care type (43%) and/or diagnosis code Z51.5 (100%). Specialist palliative care provided 1717 consultations over 507 hospital stays. Patients had 2 (median; interquartile range: 1-4) consultations during an average stay of 15.3 days (SD 15.78; median 10); the first occurred 7.0 days (SD 12.13; median 3) after admission. Of patient stays with an SPC consultation, 70% were assigned the PC Z51.5 code; 60% were referred for symptom management; 68% had cancer. One hundred forty-one patients were under a palliative specialist-either from initial hospital admission (49.6%) or later in their stay. CONCLUSIONS: Palliative care specialists provide expert input into patient management, benefitting patients and other clinicians. Administrative data inadequately capture their involvement in patient care, especially consultations, and are therefore inappropriate for reporting SPC activity. Exclusion of information related to SPC activity results in an incomplete and distorted representation of PC services and fails to acknowledge the valuable contribution made by SPC.

Inequalities in end-of-life palliative care by country of birth in New South Wales, Australia: a cohort study.

Objective This study investigated variation in in-hospital palliative care according to the decedent's country of birth. Methods A retrospective cohort study was performed of 73469 patients who died in a New South Wales public hospital between July 2010 and June 2015 and were diagnosed with a palliative care-amenable condition. Differences in receipt of palliative care by country of birth were examined using multilevel logistic regression models adjusted for confounding. Results In this cohort, 26444 decedents received palliative care during their last hospital stay. In the adjusted analysis, 40% rate differences (median odds ratio 1.39; 95% confidence interval 1.31-1.51) were observed in receipt of palliative care between country of birth groups. Conclusions There are differences in in-hospital palliative care at the end of life between population groups born in different countries living in Australia. The implementation of culturally sensitive palliative care programs may help reduce these inequalities. Further studies are needed to identify the determinants of the differences observed in this study and to investigate whether these differences persist in the community setting. What is known about the topic? International studies have reported inequities in access to palliative care between ethnic groups. What does this paper add? We observed differences in in-hospital palliative care between decedents from different countries of birth in New South Wales, Australia. These differences remained after adjusting for individual, area and hospital characteristics. What are the implications for practitioners? Implementation of culturally sensitive palliative care services and targeting groups with low rates of palliative care can reduce these inequalities and improve a patient's quality of life.

Variation in Hospital Use at the End of Life Among New South Wales Residents Who Died in Hospital or Soon After Discharge.

Objective: Hospital use increases in the last 3 months of life. We aimed to examine its association with where people live and its variation across a large health jurisdiction. Methods: We studied a number of emergency department presentations and days spent in hospital, and in-hospital deaths among decedents who were hospitalized within 30 days of death across 153 areas in New South Wales (NSW), Australia, during 2010-2015. Results: Decedents' demographics and health status were associated with hospital use. Primary care and aged care supply had no or minimal influence, as opposed to the varying effects of areal factors-socioeconomic status, remoteness, and distance to hospital last admitted. Overall, there was an approximate 20% difference in hospital use by decedents across areas. In all, 18% to 57% of areas had hospital use that differed from the average. Discussion: The observed disparity can inform targeted local efforts to strengthen the use of community care services and reduce the burden of end-of-life care on hospitals.

Variations in hospital inpatient palliative care service use: a retrospective cohort study.

OBJECTIVE: Use of palliative care in hospitals for people at end of life varies. We examined rate and time of in-hospital palliative care use and associated interhospital variations. METHODS: We used admissions from all hospitals in New South Wales, Australia, within a 12-month period, for a cohort of adults who died in 73 public acute care hospitals between July 2010 and June 2014. Receiving palliative care and its timing were based on recorded use. RESULTS: Among 90 696 adults who died, 27% received palliative care, and the care was initiated 7.6 days (mean; SD: 3.3 days) before death. Over the 5-year period, the palliative care rate rose by 58%, varying extent across chronic conditions. The duration of palliative care before death declined by 7%. Patient (demographics, morbidities and service use) and hospital factors (size, location and availability of palliative care unit) explained half of the interhospital variation in outcomes: adjusted IQR in rate and duration of palliative care among hospitals were 23%-39% and 5.2-8.7 days, respectively. Hospitals with higher rates often initiated palliative care earlier (correlation: 0.39; p<0.01). CONCLUSION: Despite an increase over time in the palliative care rate, its initiation was late and of brief duration. Palliative care use was associated with patient and hospital characteristics; however, half of the between hospital variation remained unexplained. The observed suboptimal practices and variability indicate the need for expanded and standardised use of palliative care supported by assessment tools, service enhancement and protocols.

Inpatient palliative care of people dying in New South Wales hospitals or soon after discharge.

BACKGROUND: Palliative care can benefit all patients with life-limiting diseases. AIM: To describe hospital use in the final year of life, timing of palliative care and variations by age and disease for patients receiving inpatient palliative care. METHODS: Retrospective cohort study of all New South Wales residents aged 50 years and older who died (decedents) between July 2010 and June 2015 in hospital or within 30 days of discharge. Care type and diagnosis codes identified decedents who received inpatient palliative care. RESULTS: Of 150 770 decedents, 34.4% received palliative care a median of 10 days before death. Decedents were more likely to receive palliative care if they had cancer (64.7% vs 13.3% for those without chronic conditions) or were younger (46.3% vs 25.0% of the oldest decedents). In their last year of life, palliated decedents, on average, had three emergency department presentations and four hospital admissions - one involving surgery and one where palliation was the intent of care. Of the 30.1 days spent in hospital, 8.7 days involved palliative care. Older age and non-cancer diagnoses were associated with fewer days of inpatient palliation and shorter time between first palliative admission and death. Decedents dying out of hospital started palliative care 18 days earlier than those dying in hospital. CONCLUSION: Most decedents did not receive palliative care during hospital admission, and even then only very late in life, limiting its benefits. Improved recognition of palliative need, including earlier identification regardless of age and disease, will enhance the quality of care for the dying.

Accuracy of inter-hospital transfer information in Australian hospital administrative databases.

Inter-hospital transfers improve care delivery for which sending and receiving hospitals both accountable for patient outcomes. We aim to measure accuracy in recorded patient transfer information (indication of transfer and hospital identifier) over 2 years across 121 acute hospitals in New South Wales, Australia. Accuracy rate for 127,406 transfer-out separations was 87 per cent, with a low variability across hospitals (10% differences); it was 65 per cent for 151,978 transfer-in admissions with a greater inter-hospital variation (36% differences). Accuracy rate varied by departure and arrival pathways; at receiving hospitals, it was lower for transfer-in admission via emergency department (incidence rate ratio = 0.52, 95% confidence interval: 0.51-0.53) versus direct admission. Transfer-out data were more accurate for transfers to smaller hospitals (incidence rate ratio = 1.06, 95% confidence interval: 1.03-1.08) or re-transfers (incidence rate ratio > 1.08). Incorporation of transfer data from sending and receiving hospitals at patient level in administrative datasets and standardisation of documentation across hospitals would enhance accuracy and support improved attribution of hospital performance measures.

Variation in out-of-hospital death among palliative care inpatients across public hospitals in New South Wales, Australia.

BACKGROUND: Place of death is an important indicator in palliative care, as out-of-hospital death is often preferred by patients and is less costly for the healthcare system. AIM: To examine variation and contributing factors in out-of-hospital death after receiving palliative care in hospital to inform improvement in transition of care between hospitals and communities. METHODS: Using hospital linked data (July 2010, June 2015) we followed individuals aged 50 or older who received palliative care in hospital and within 3 months to death who were last admitted to a public acute-care hospital in New South Wales, Australia (73 hospitals). RESULTS: Among 25 359 palliative care inpatients, 3677 (14%) died out of hospital. The out-of-hospital death rate was lower for younger patients, males and those living in the most deprived areas; it was higher for cancer patients and those who received palliative care before their last admission. Hospital size, location and availability of hospice care unit were not influential. Across hospitals, the median crude rate of out-of-hospital death was 14% (interquartile range 10-19%). The contributing factors explained 19% of the variation, resulting in a rate difference of 44% between hospitals with high versus low rates; 25% of hospitals had a higher and 14% had a lower than average adjusted out-of-hospital death rate. CONCLUSION: The majority of patients who received palliative care in hospital stayed in hospital until death. The variation in out-of-hospital death across hospitals was considerable and mostly remained unexplained. This variability warrants investigation into transition of palliative care between hospitals and communities to inform interventions.

Variation in the incidence and timing of diagnosis of hospital-associated venous thromboembolism using linked administrative data.

Venous thromboembolism (VTE) is a potentially preventable adverse effect of hospitalisation. Inter-hospital variation in the incidence of hospital-associated VTE (HA-VTE) and timing of diagnosis (in-hospital or post-discharge) in New South Wales public hospitals were examined. Large variations in incidence (22% risk difference) and post-discharge diagnosis (115% odds difference) were evident after adjustment for case mix, which only explained 59% and 32% of inter-hospital variation respectively. The need for improved compliance with best practice guidelines is reinforced.