Finite mixture regression model with random effects: application to neonatal hospital length of stay

A two-component mixture regression model that allows simultaneously for heterogeneity and dependency among observations is proposed. By specifying random effects explicitly in the linear predictor of the mixture probability and the mixture components, parameter estimation is achieved by maximising the corresponding best linear unbiased prediction type log-likelihood. Approximate residual maximum likelihood estimates are obtained via an EM algorithm in the manner of generalised linear mixed model (GLMM). The method can be extended to a g-component mixture regression model with the component density from the exponential family, leading to the development of the class of finite mixture GLMM. For illustration, the method is applied to analyse neonatal length of stay (LOS). It is shown that identification of pertinent factors that influence hospital LOS can provide important information for health care planning and resource allocation. (C) 2002 Elsevier Science B.V. All rights reserved.

Modelling inpatient length of stay by a hierarchical mixture regression via the EM algorithm

The modelling of inpatient length of stay (LOS) has important implications in health care studies. Finite mixture distributions are usually used to model the heterogeneous LOS distribution, due to a certain proportion of patients sustaining-a longer stay. However, the morbidity data are collected from hospitals, observations clustered within the same hospital are often correlated. The generalized linear mixed model approach is adopted to accommodate the inherent correlation via unobservable random effects. An EM algorithm is developed to obtain residual maximum quasi-likelihood estimation. The proposed hierarchical mixture regression approach enables the identification and assessment of factors influencing the long-stay proportion and the LOS for the long-stay patient subgroup. A neonatal LOS data set is used for illustration, (C) 2003 Elsevier Science Ltd. All rights reserved.

Type of housing predicts rate of readmission to hospital but not length of stay in people with schizophrenia on the Gold Coast in Queensland

Accommodation is considered to be important by institutions interested in mental health care both in Australia and internationally. Some authorities assert that no component of a community mental health system is more important than decent affordable housing. Unfortunately there has been little research in Australia into the consequences of discharging people with a primary diagnosis of schizophrenia to different types of accommodation. This paper uses archival data to investigate the outcomes for people with schizophrenia discharged to two types of accommodation. The types of accommodation chosen are the person's own home and for-profit boarding house. These two were chosen because the literature suggests that they are respectively the most and least desirable types of accommodation. Results suggest that people with schizophrenia who were discharged to boarding houses are significantly more likely to be readmitted to the psychiatric unit of Gold Coast Hospital although their length of stay in hospital is not significantly different. (author abstract)

Length of Stay, Discharge Destination, and Functional Improvement: Utility of the Australian National Subacute and Nonacute Patient Casemix Classification

Background and Purpose - Although implemented in 1998, no research has examined how well the Australian National Subacute and Nonacute Patient (AN-SNAP) Casemix Classification predicts length of stay (LOS), discharge destination, and functional improvement in public hospital stroke rehabilitation units in Australia. Methods - 406 consecutive admissions to 3 stroke rehabilitation units in Queensland, Australia were studied. Sociode-mographic, clinical, and functional data were collected. General linear modeling and logistic regression were used to assess the ability of AN-SNAP to predict outcomes. Results - AN-SNAP significantly predicted each outcome. There were clear relationships between the outcomes of longer LOS, poorer functional improvement and discharge into care, and the AN-SNAP classes that reflected poorer functional ability and older age. Other predictors included living situation, acute LOS, comorbidity, and stroke type. Conclusions - AN-SNAP is a consistent predictor of LOS, functional change and discharge destination, and has utility in assisting clinicians to set rehabilitation goals and plan discharge.

An Australian comparison of specialist care of acute myocardial infarction

Objective. To determine whether patients hospitalized with acute myocardial infarction (AMI) in an Australian setting receive better pharmacological care if managed by cardiologists than by non-cardiologists. Design. Retrospective chart review of patients hospitalized between 1 January 1997 and 30 June 1998, undertaken by abstractors blind to study objectives. Setting. One tertiary and two community hospitals in south-east Queensland, Australia, in which all patients admitted with AMI were cared for by cardiologists and general physicians, respectively. Study participants. Two cohorts of consecutive patients satisfying diagnostic criteria for AMI: 184 in the tertiary hospital and 207 in the community hospitals. Main outcome measures. Frequency of use, in highly eligible patients, of thrombolysis, P-blockers, aspirin, angiotensin-converting enzyme (ACE) inhibitors, lipid-lowering agents, nitrates, and calcium antagonists. Cohorts were compared for differences in prognostic factors or illness severity. Results. In community hospital patients, there was greater use of thrombolysis [100% versus 83% in the tertiary hospital; difference 17%, 95% confidence interval (CI) 11-26%; P < 0.001] and of ACE inhibitors (84% versus 66%; difference 18%, 95% CI 3-34%; P = 0.02), and lower median length of stay (6.0 days versus 7.0 days; P = 0.001) compared with tertiary hospital patients. Frequency of use of other drugs, and adjusted rates of death and re-infarction were the same for both cohorts. Conclusions. With respect to pharmacological management of patients hospitalized with AMI, cardiologists and general physicians appear to provide care of similar quality and achieve equivalent outcomes. Further studies are required to confirm the generalizability of these results to Australian practice as a whole.

Trends in the use of hospital beds by older people in Australia: 1993-2002

Objective: To determine trends in use of Australian acute hospital inpatient services by older patients. Design and data sources: Secondary analysis of hospital data from the Australian Institute of Health and Welfare in the period 1993-94 to 2001-02, with population data for this period from the Australian Bureau of Statistics. Outcome measures: Population-based rates of hospital separations and bed utilisation. Results: The Australian aged population (65 years and older) increased by 18% compared with total population growth of 10%, yet the proportion of hospital beds occupied by older patients remained stable at 47%. The most substantial changes were observed in the population aged 75 years and older, with separations increasing by 89%, length of stay reducing by 35% and bed utilisation increasing by 23%. However, rates of bed utilisation (in relation to population) declined among older groups (10% decline in per capita use in population 75 years and older), but increased in the younger population (1% increase in per capita use in people younger than 65 years). Conclusion: Important trends in use of inpatient services were identified in this study. These trends are contrary to common perception. Ageing of the Australian population was not associated with an increase in the proportion of hospital beds used by older patients.

A prospective multicentre study of pharmacist initiated changes to drug therapy and patient management in acute care government funded hospitals

Aims To determine the cost savings of pharmacist initiated changes to hospitalized patients' drug therapy or management in eight major acute care government funded teaching hospitals in Australia. Methods This was a prospective study performed in eight hospitals examining resource implications of pharmacists' interventions assessed by an independent clinical panel. Pharmacists providing clinical services to inpatients recorded details of interventions, defined as any action that directly resulted in a change to patient management or therapy. An independent clinical review panel, convened at each participating centre, confirmed or rejected the clinical pharmacist's assessment of the impact on length of stay (LOS), readmission probability, medical procedures and laboratory monitoring and quantified the resultant changes, which were then costed. Results A total of 1399 interventions were documented. Eight hundred and thirty-five interventions impacted on drug costs alone. Five hundred and eleven interventions were evaluated by the independent panels with three quarters of these confirmed as having an impact on one or more of: length of stay, readmission probability, medical procedures or laboratory monitoring. There were 96 interventions deemed by the independent panels to have reduced LOS and 156 reduced the potential for readmission. The calculated savings was $263 221 for the eight hospitals during the period of the study. This included $150 307 for length of stay reduction, $111 848 for readmission reduction. Conclusions The annualized cost savings relating to length of stay, readmission, drugs, medical procedures and laboratory monitoring as a result of clinical pharmacist initiated changes to hospitalized patient management or therapy was $4 444 794 for eight major acute care government funded teaching hospitals in Australia.

The use of cusum analysis in the early detection and management of hospital bed occupancy crises

Objective: To assess the value of cusum analysis in hospital bed management. Design: Comparative analysis of medical patient flows, bed occupancy, and emergency department admission rates and access block over 2 years. Setting: Internal Medicine Services and Emergency Department in a teaching hospital. Interventions: Improvements in bed use and changes in the level of available beds. Main outcome measures: Average length of stay; percentage occupancy of available beds; number of patients waiting more than 8 hours for admission (access block); number of medical patients occupying beds in non-medical wards; and number of elective surgical admissions. Results: Cusum analysis provided a simple means of revealing important trends in patient flows that were not obvious in conventional time-series data. This prompted improvements in bed use that resulted in a decrease of 9500 occupied bed-days over a year. Unfortunately and unexpectedly, after some initial improvement, the levels of access block, medical ward congestion and elective surgical admissions all then deteriorated significantly. This was probably caused by excessive bed closures in response to the initial improvement in bed use. Conclusion: Cusum analysis is a useful technique for the early detection of significant changes in patient flows and bed use, and in determining the appropriate number of beds required for a given rate of patient flow.

Empiric management of community-acquired pneumonia in Australian emergency departments

Objective: To describe empiric community-acquired pneumonia (CAP) management in Australian hospital emergency departments (EDs) and evaluate this against national guidelines, including use of the pneumonia severity index and antibiotic selection. Design: A multicentre, cross-sectional, retrospective audit, April 2003 to February 2005. Setting: 37 Australian hospitals: 22 principal referral hospitals, six large major city hospitals, four large regional hospitals, four medium hospitals and one private hospital. Participants: Adult patients with a diagnosis of CAP made in the ED. Data on 20 consecutive CAP ED presentations were collected in participating hospitals. Outcome measures: Documented use of the pneumonia severity index, initial antibiotic therapy prescribed in the ED, average length of stay, inpatient mortality, and concordance with national guidelines. Results: 691 CAP presentations were included. Pneumonia severity index use was documented in 5% of cases. Antibiotic therapy covering common bacterial causes of CAP was prescribed in 67% of presentations, although overall concordance with national guidelines was 18%. Antibiotic prescribing was discordant due to inadequate empiric antimicrobial cover, allergy status (including contraindication to penicillin), inappropriate route of administration and/or inappropriate antibiotic choice according to recommendations. There was no significant difference between concordant and discordant antibiotic prescribing episodes in average length of stay (5.0 v 5.7 days; P=0.22) or inpatient mortality (1.6% v 4.1%; chi(2) = 1.82; P=0.18). Conclusions: Antibiotic therapy for CAP prescribed in Australian EDs varied. Concordance with national CAP guidelines was generally low. Targeted interventions are required to improve concordance.

The costs of caring for patients in a tertiary referral Australian intensive care unit

We determined the direct cost of an Intensive Care Unit (ICU) bed in a tertiary referral Australian ICU and the cost drivers thereof, by retrospectively analysing a number of prospectively designed Hospital- and Unit-specific electronic databases. The study period was a financial year, from 1 July 2002 to 30 June 2003. There were 1615 patients occupying 5692 fractional occupied bed days at a total cost of A$15,915,964, with an average length of stay of 3.69 days (range 0.5-77, median 1.06, interquartile range 2.33). The main cost driver not incorporated into this analysis was blood products (paid for centrally). The average costs of an ICU day and total stay per patient were A$2670 and A$9852 respectively. Staff-related charges were 68.76%, with consumables related expenditure making up 19.65%, clinical support services 9.55% and capital equipment 2.04%. Overtime charges and nursing agency staff were 19.4% of staff-related charges (2.9% for agency staff), 3.9% lower than expenditure associated with full-time employment charges, such as pension and leave. The emergency nature of ICU means it is difficult to accurately set a nursing establishment to cater for all admissions and therefore it is hard to decide what is an acceptable percentage difference between agency/overtime costs compared with the costs associated with full-time staff appointments. Consumable expenditure is likely to increase the most with new innovation and therapies. Using protocol driven practices may tighten and control costs incurred in ICU.

The effect of an algorithm-based sedation guideline on the duration of mechanical ventilation in an Australian intensive care unit

To examine the effect of an algorithm-based sedation guideline developed in a North American intensive care unit (ICU) on the duration of mechanical ventilation of patients in an Australian ICU. The intervention was tested in a pre-intervention, post-intervention comparative investigation in a 14-bed adult intensive care unit. Adult mechanically ventilated patients were selected consecutively (n =322) The pre-intervention and post-intervention groups were similar except for a higher number of patients with a neurological diagnosis in the pre-intervention group. An algorithm-based sedation guideline including a sedation scale was introduced using a multifaceted implementation strategy. The median duration of ventilation was 5.6 days in the post-intervention group, compared with 4.8 days for the pre-intervention group (P = 0.99). The length of stay was 8.2 days in the post-intervention group versus 7.1 days in the pre-intervention group (P = 0.04). There were no statistically significant differences for the other secondary outcomes, including the score on the Experience of Treatment in ICU 7 item questionnaire, number of tracheostomies and number of self-extubations. Records of compliance to recording the sedation score during both phases revealed that patients were slightly more deeply sedated when the guideline was used. The use of the algorithm-based sedation guideline did not reduce duration of mechanical ventilation in the setting of this study.

Controlled trial of multidisciplinary care teams for acutely ill medical inpatients: enhanced multidisciplinary care

Background: Acute hospital general medicine services care for ageing complex patients, using the skills of a range of health-care providers. Evidence suggests that comprehensive early assessment and discharge planning may improve efficiency and outcomes of care in older medical patients. Aim: To enhance assessment, communication, care and discharge planning by restructuring consistent, patient-centred multidisciplinary teams in a general medicine service. Methods: Prospective controlled trial enrolling 1538 consecutive medical inpatients. Intervention units with additional allied health staff formed consistent multidisciplinary teams aligned with inpatient admitting units rather than wards; implemented improved communication processes for early information collection and sharing between disciplines; and specified shared explicit discharge goals. Control units continued traditional, referral-based multidisciplinary models with existing staffing levels. Results: Access to allied health services was significantly enhanced. There was a trend to reduced index length of stay in the intervention units (7.3 days vs 7.8 days in control units, P = 0.18), with no change in 6-month readmissions. in-hospital mortality was reduced from 6.4 to 3.9% (P = 0.03); less patients experienced functional decline in hospital (P = 0.04) and patients' ratings of health status improved (P = 0.02). Additional staffing costs were balanced by potential bed-day savings. Conclusion: This model of enhanced multidisciplinary inpatient care has provided sustainable efficiency gains for the hospital and improved patient outcomes.

Acupressure for the in-patient treatment of nausea and vomiting in early pregnancy: A randomized control trial

Objective: The purpose of this study was to evaluate the efficacy of acupressure at the P6 point for the in-patient treatment of severe nausea and vomiting in early pregnancy. Study design: This was a prospective single-blind randomized control trial that involved 80 patients with nausea and vomiting plus ketonuria before 14 weeks of gestation. Results: There was no difference between length of stay, amount of medication, or fluid required between the acupressure and placebo groups, although acupressure reduced the number of patients who stayed >= 4 nights in the hospital. Acupressure was well tolerated and not associated with an increase in perinatal morbidity or death. Conclusion: The use of acupressure Lit the P6 point does not reduce the amount of antiemetric medication that is required, the requirement for intravenous fluid, and median duration of inpatient stay more than the use of placebo. A small reduction was seen in the number of women who required >= 4 days in the hospital. (c) 2006 Mosby, Inc. All rights reserved.

An Incremental EM-based Learning Approach for On-Line Prediction of Hospital Resource Utilization

Objective: Inpatient length of stay (LOS) is an important measure of hospital activity, health care resource consumption, and patient acuity. This research work aims at developing an incremental expectation maximization (EM) based learning approach on mixture of experts (ME) system for on-line prediction of LOS. The use of a batchmode learning process in most existing artificial neural networks to predict LOS is unrealistic, as the data become available over time and their pattern change dynamically. In contrast, an on-line process is capable of providing an output whenever a new datum becomes available. This on-the-spot information is therefore more useful and practical for making decisions, especially when one deals with a tremendous amount of data. Methods and material: The proposed approach is illustrated using a real example of gastroenteritis LOS data. The data set was extracted from a retrospective cohort study on all infants born in 1995-1997 and their subsequent admissions for gastroenteritis. The total number of admissions in this data set was n = 692. Linked hospitalization records of the cohort were retrieved retrospectively to derive the outcome measure, patient demographics, and associated co-morbidities information. A comparative study of the incremental learning and the batch-mode learning algorithms is considered. The performances of the learning algorithms are compared based on the mean absolute difference (MAD) between the predictions and the actual LOS, and the proportion of predictions with MAD < 1 day (Prop(MAD < 1)). The significance of the comparison is assessed through a regression analysis. Results: The incremental learning algorithm provides better on-line prediction of LOS when the system has gained sufficient training from more examples (MAD = 1.77 days and Prop(MAD < 1) = 54.3%), compared to that using the batch-mode learning. The regression analysis indicates a significant decrease of MAD (p-value = 0.063) and a significant (p-value = 0.044) increase of Prop(MAD