Dennis Petrie

OBJECTIVE To develop a patient-level simulation model for predicting lifetime health outcomes of patients with type 1 diabetes and as a tool for economic evaluation of type 1 diabetes treatment based on data from a large, longitudinal cohort. RESEARCH DESIGN AND METHODS Data for model development were obtained from the Swedish National Diabetes Register. We derived parametric proportional hazards models predicting absolute risk of diabetes complications and death based on a wide range of clinical variables and history of complications. We used linear regression models to predict risk factor progression. Internal validation was performed, estimates of life expectancies for different age-sex strata were computed, and the impact of key risk factors on life expectancy was assessed. RESULTS The study population consisted of 27,841 patients with type 1 diabetes with a mean duration of follow-up of 7 years. Internal validation showed good agreement between predicted and observed cumulative...

To test the effectiveness of a theoretically based text-message intervention to reduce binge drinking among socially disadvantaged men. A multi-centre parallel group, pragmatic, individually randomized controlled trial. Community-based study conducted in four regions of Scotland. A total of 825 men aged 25-44 years recruited from socially disadvantaged areas who had two or more episodes of binge drinking (> 8 UK units on a single occasion) in the preceding 28 days: 411 men were randomized to the intervention and 414 to the control. A series of 112 interactive text messages was delivered by mobile phone during a 12-week period. The intervention was structured around the Health Action Process Approach, a comprehensive model which allows integration of a range of evidence-based behaviour change techniques. The control group received 89 texts on general health, with no mention of alcohol or use of behaviour change techniques. The primary outcome measure was the proportion of men cons...

To compare the determinants of initial statin prescribing between New Zealand and Australia. New Zealand has a system-wide absolute risk-based approach to primary care cardiovascular disease (CVD) management, while Australia has multiple guidelines. Classification and Regression Tree (CART) analysis of two observational studies of primary care CVD management from New Zealand (PREDICT-CVD) and Australia (AusHeart). Over 80% of eligible New Zealanders have been screened for CVD risk. PREDICT-CVD is used by approximately one-third of New Zealand GPs to perform web-based CVD risk assessment in routine practice, with the sample consisting of 126,519 individuals risk assessed between 1 January 2007 and 30 June 2014. AusHeart is a cluster-stratified survey of primary care CVD management that enrolled 534 GPs from across Australia, who in turn recruited 1381 patients between 1 April and 30 June 2008. Eligibility was restricted to 55-74year old patients without prior CVD. The CART analyses d...

Underemployment (defined as when a person in paid employment works for fewer hours than their desired full working capacity) is increasingly recognised as a component of employment precarity. This paper sought to investigate the effects of underemployment on the mental health of people with disabilities. Using 14 waves of the Household, Income and Labour Dynamics in Australia survey, we used fixed-effects models to assess whether the presence of a disability modified the association between underemployment and mental health. Both disability and underemployment were assessed as time-varying factors. Measures of effect measure modification were presented on the additive scale. The experience of underemployment was associated with a significantly greater decline in mental health when a person reported a disability (mean difference -1.38, 95% CI -2.20 to -0.57) compared with when they did not report a disability (mean difference -0.49, 95% CI -0.84 to -0.14). The combined effect of bein...

Status inconsistency refers to a discrepancy between the position a person holds in one domain of their social environment comparative to their position in another domain. For example, the experience of being overeducated for a job, or not using your skills in your job. We sought to assess the relationship between status inconsistency and mental health using 14 annual waves of cohort data. We used two approaches to measuring status inconsistency: 1) being overeducated for your job (objective measure); and b) not using your skills in your job (subjective measure). We implemented a number of methodological approaches to assess the robustness of our findings, including instrumental variable, random effects, and fixed effects analysis. Mental health was assessed using the Mental Health Inventory-5. The random effects analysis indicates that only the subjective measure of status inconsistency was associated with a slight decrease in mental health (β-1.57, 95% -1.78 to -1.36, p < 0.001). This size of these coefficients was maintained in the instrumental variable analysis. We suggest that status inconsistency might explain some of the relationship between social determinants (such as work and education) and health outcomes.

Many treatments are evaluated using quasi-experimental pre-post studies susceptible to regression to the mean (RTM). Ignoring RTM could bias the economic evaluation. We investigated this issue using the contemporary example of total knee replacement (TKR), a common treatment for end-stage osteoarthritis of the knee. Data (n = 4796) were obtained from the Osteoarthritis Initiative database, a longitudinal observational study of osteoarthritis. TKR patients (n = 184) were matched to non-TKR patients, using propensity score matching on the predicted hazard of TKR and exact matching on osteoarthritis severity and health-related quality of life (HrQoL). The economic evaluation using the matched control group was compared to the standard method of using the pre-surgery score as the control. Matched controls were identified for 56% of the primary TKRs. The matched control HrQoL trajectory showed evidence of RTM accounting for a third of the estimated QALY gains from surgery using the pre-s...

This study investigated the extent that psychosocial job stressors had lasting effects on a scaled measure of mental health. We applied econometric approaches to a longitudinal cohort to: (1) control for unmeasured individual effects; (2) assess the role of prior (lagged) exposures of job stressors on mental health and (3) the persistence of mental health. We used a panel study with 13 annual waves and applied fixed-effects, first-difference and fixed-effects Arellano-Bond models. The Short Form 36 (SF-36) Mental Health Component Summary score was the outcome variable and the key exposures included: job control, job demands, job insecurity and fairness of pay. Results from the Arellano-Bond models suggest that greater fairness of pay (β-coefficient 0.34, 95% CI 0.23 to 0.45), job control (β-coefficient 0.15, 95% CI 0.10 to 0.20) and job security (β-coefficient 0.37, 95% CI 0.32 to 0.42) were contemporaneously associated with better mental health. Similar results were found for the fixed-effects and first-difference models. The Arellano-Bond model also showed persistent effects of individual mental health, whereby individuals' previous reports of mental health were related to their reporting in subsequent waves. The estimated long-run impact of job demands on mental health increased after accounting for time-related dynamics, while there were more minimal impacts for the other job stressor variables. Our results showed that the majority of the effects of psychosocial job stressors on a scaled measure of mental health are contemporaneous except for job demands where accounting for the lagged dynamics was important.

Objective The aim of the present study was to quantify and understand the utilisation of linked hospital data for research purposes across Australia over the past two decades. Methods A systematic review was undertaken guided by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2009 checklist. Medline OVID, PsycINFO, Embase, EconLit and Scopus were searched to identify articles published from 1946 to December 2014. Information on publication year, state(s) involved, type of data linkage, disease area and purpose was extracted. Results The search identified 3314 articles, of which 606 were included; these generated 629 records of hospital data linkage use across all Australian states and territories. The major contributions were from Western Australia (WA; 51%) and New South Wales (NSW; 32%) with the remaining states and territories having significantly fewer publications (total contribution only 17%). WA’s contribution resulted from a steady increase fr...

We investigated the predictors of long-term gains in quality-adjusted life years (QALYs) from total knee arthroplasty (TKA) and the patient attributes that predicted cost-effective TKA. Data on TKA patients (n = 570) from 2006 to 2007 were extracted from a single-institution registry. QALY gains over 7 years post surgery were calculated from health-related quality of life (HrQoL) scores measured preoperatively and annually postoperatively using the short-form health survey (SF-12) instrument. Multivariate linear regression analysis investigated the predictors of QALY gain from TKA from a broad range of preoperative patient characteristics and was used to predict QALY gains for each individual. Patients were grouped into deciles according to their predicted QALY gain, and the cost-effectiveness of each decile was plotted on the cost-effectiveness plane. Patient attribute differences between deciles were decomposed. After exclusions and dropout, data were available for 488 patients. The average estimated QALY gain over 7 years was 0.77 (95% confidence interval [CI] 0.70-0.83). Predictors significantly associated with smaller QALY gains were comorbidities (Charlson comorbidity index 3+ coefficient -0.54 CI -0.15 to -0.92), the absence of severe osteoarthritis in the ipsilateral knee (-0.51 CI -0.16 to -0.85), preoperative HrQoL (standardized coefficient -0.34 CI -0.26 to -0.43), the requirement for an interpreter (-0.24 CI -0.05 to -0.44), and age (-0.01 CI -0.01 to -0.02). The largest difference between cost-effective and non-cost-effective deciles was relatively high preoperative HrQoL in the non-cost-effective decile. TKA is likely to be cost-effective for most patients except those with unusually high preoperative HrQoL or a lack of severe osteoarthritis. The poorer outcomes for those requiring an interpreter requires further research.

ABSTRACT Alcohol is a complex good which is ingrained within the social fabric of Australian culture. This is even more apparent within rural communities which are considered to have higher levels of risky alcohol consumption. While alcohol provides pleasure to many, it has devastating effects for others. Not only are individuals themselves affected, but also their families and the community at large. This thesis considers a number of the economic aspects associated with alcohol consumption in rural Australia. This thesis was completed in conjunction with a larger ongoing project: Alcohol Action in Rural Communities (AARC), which is a cost-benefit analysis of community-wide interventions to reduce alcohol-related harm in rural Australia. Given the study involves 20 rural communities in New South Wales, Australia; much of the analysis is centred around data collected from these communities. This thesis examines the amount and patterns of alcohol consumption, some of the associated harms, the value of reducing these harms, plus professionals’ and the public’s opinions on interventions, viewed as an optimal solution to reduce alcohol-related harm for the communities within this study. First, those factors which affect individuals’ decisions regarding how much alcohol to consume are examined through econometric modelling of the demand for alcohol. Moreover, a theoretical model is derived whereby individuals choose both the intensity and frequency of alcohol consumption in order to maximise their utility. This is then used to examine those factors that affect the relationship between intensity and frequency of alcohol consumption for individuals within the 20 communities of the study. Secondly, the impact that this alcohol consumption has in terms of the effect on morbidity, crime and traffic accidents within these 20 communities is analysed. The relationship between selfreported quality of life using the EQ5D (a quality of life instrument) and self-reported risky alcohol use are examined using ordered Probit and Tobit models. Also, the relationship between community levels of risky drinking and crime and traffic accidents that occur in alcohol-related times is analysed, controlling for the underlying level of crime in the community by using the rate of incidents that occur in non-alcohol-related times. It is found that rural communities in Australia are experiencing a sizeable amount of potentially avoidable harm due to risky alcohol use. Thirdly, the value of reducing alcohol-related harm in these communities, in terms of the amount households are willing to pay for these reductions, is estimated using contingent valuation methods. Individuals’ willingness to pay for a percentage reduction in alcohol-related harm is estimated using both a postal questionnaire for the 20 communities, plus a face-to-face questionnaire conducted in two of these communities. The face-to-face questionnaire is also used to conduct a double-bounded dichotomous choice experiment, to investigate the willingness of households to pay for a reduction in a number of different types of alcohol-related harm. Finally, the views on the optimal policy options to reduce alcohol-related harm for rural communities in Australia are examined for both professionals and the general public. A sample of drug and alcohol (D&A) professionals were asked to allocate a budget of $100,000 to a number of interventions in order to reduce alcohol-related harm in a hypothetical rural community. The D&A professionals most commonly selected interventions include training general practitioners (GPs), targeting high-risk groups, developing a harm-reduction code of practice, expanding social work services and the training of emergency department staff. Additionally, individuals from the general public were asked via a postal questionnaire to allocate a percentage of total funds to eight intervention areas in order to reduce alcohol-related harm in their communities. The top three intervention areas given the most funding, on average, by the public were school-based interventions, educational messages in the media, and greater police enforcement. There is no doubt that alcohol consumption causes a substantial amount of harm for rural communities within Australia. This thesis has provided information to inform the development of interventions tailored to specific communities and has derived estimates which can be used to help evaluate the cost-benefit of these interventions. There is still additional research to be done in order to obtain more accurate estimates of the exact effect alcohol has on rural communities and thus comprehensively evaluate which interventions are likely to be the most cost-effective in reducing this harm.

People with type 1 diabetes have reduced life expectancy (LE) compared with the general population. Our aim is to quantify mortality changes from 2002 to 2011 in people with type 1 diabetes in Sweden. This study uses health records from the Swedish National Diabetes Register (NDR) linked with death records. Abridged period life tables for those with type 1 diabetes aged 20 years and older were derived for 2002-06 and 2007-11 using Chiang's method. Cox proportional hazard models were used to assess trends in overall and cause-specific mortality. There were 27,841 persons aged 20 years and older identified in the NDR as living with type 1 diabetes between 2002 and 2011, contributing 194,685 person-years of follow-up and 2,018 deaths. For men with type 1 diabetes, the remaining LE at age 20 increased significantly from 47.7 (95% CI 46.6, 48.9) in 2002-06 to 49.7 years (95% CI 48.9, 50.6) in 2007-11. For women with type 1 diabetes there was no significant change, with an LE at age 2...

This paper develops a theoretical model for the demand of alcohol where intensity and frequency of consumption are separate choices made by individuals in order to maximize their utility. While distinguishing between intensity and frequency of consumption may be unimportant for many goods, this is clearly not the case with alcohol where the likelihood of harm depends not only on the total amount consumed but also on the pattern of use. The results from the theoretical model are applied to data from rural Australia in order to investigate the factors that affect the patterns of alcohol use for this population group. This research can play an important role in informing policies by identifying those factors which influence preferences for patterns of risky alcohol use and those groups and communities who are most at risk of harm.

Life expectancy at birth is the most commonly used measure for health system output. However, there are a number of reasons why it may be a poor proxy. First, life expectancy assumes a stationary population and thus does not take into account the current demographic structure of a country; and second, the output of a health system should be measured in terms of the value-added to the population’s health status rather than health status itself. The paper develops a new measure of health system output, the Incremental Life Years to address these problems. The new measure is applied to study health system output, efficiency and total factor productivity in OECD countries for the years 2000 and 2004. The new measure provides different results compared to those based on the traditional life expectancy measure, and the differences are further accentuated when changes in efficiency and productivity are estimated.

Alcohol misuse represents one of the leading causes of preventable death, illness and injury in Australia. Extensive research exists estimating the effect of risky alcohol use on mortality but little research quantifies the impact of risky alcohol consumption on morbidity. Estimates are needed to measure the benefit of interventions which reduce risky alcohol use. Ordered probit and tobit models are used to analyse the impact of risky drinking on self reported health status using data from rural Australians. It is found that risky alcohol use is associated with lower self-reported health however the average effect is small apart for those drinking at very high risk.

To determine if taxation policies that increase the price of alcohol differentially reduce alcohol consumption for heavy drinkers in Australia. A two-part demand model for alcohol consumption is used to determine the price elasticity of alcohol. Quantile regression is used to determine the price elasticity estimates for various levels of consumption. The study uses Australian data collected by the National Drug Strategy Household Survey for the years 2001, 2004 and 2007. Measures of individual annual alcohol consumption were derived from three waves of the National Drug Strategy Household Survey; alcohol prices were taken from market research reports. For the overall population of drinkers, a 1 % increase in the price of alcohol was associated with a 0.96 % (95 % CI -0.35 %, -1.57 %) reduction in alcohol consumption. For those in the highest 10 % of drinkers by average amount consumed, a 1 % increase in the price of alcohol was associated with a 1.26 % (95 % CI 0.82 %, 1.70 %) reduction in consumption. Within Australia, policies that increase the price of alcohol are about equally effective in relative terms for reducing alcohol consumption both for the general population and among those who drink heavily.

This paper assesses whether the concession card, which offers discounted out-of-pocket costs for prescription medicines in Australia, affects discontinuation and adherence to statin therapy. The analysis uses data from the Australian Hypertension and Absolute Risk Study (AusHEART), which involves patients aged 55 years and over who visited a GP between April and June 2008. Socioeconomic and clinical information was collected and linked to administrative data on pharmaceutical use. Patients without a concession card were 63% more likely (hazard ratio (HR) 95% confidence interval (CI): 1.14-2.33) to discontinue and 60% (odds ratio (OR) CI: 1.04-2.44) more likely to fail to adhere to therapy compared to concessional patients. Smokers were 2.12 (HR CI: 1.39-3.22) times more likely to discontinue use and 2.23 (OR CI: 1.35-3.71) times more likely to fail to adhere compared to non-smokers. Patients who had recently initiated statin medication were also 2.28 (HR CI: 1.22-4.28) times more likely to discontinue use. In conclusion, higher copayments act as a disincentive for persistent and adherent use of statin medication.

Alcohol misuse represents one of the leading causes of preventable death, illness and injury in Australia. Extensive research exists estimating the effect of risky alcohol use on mortality, but little research quantifies the impact of risky alcohol consumption on morbidity. Such estimates are required to more precisely measure the benefit of interventions aimed at reducing risky alcohol use. Ordered probit

Socially disadvantaged men are at a substantially higher risk of developing alcohol-related problems. The frequency of heavy drinking in a single session is high among disadvantaged men. Brief alcohol interventions were developed for, and are usually delivered in, healthcare settings. The group who binge drink most frequently, young to middle-aged disadvantaged men, have less contact with health services and there is a need for an alternative method of intervention delivery. Text messaging has been used successfully to modify other adverse health behaviours. This study will test whether text messages can reduce the frequency of binge drinking by disadvantaged men. Disadvantaged men aged 25 to 44 years who drank >8 units of alcohol at least twice in the preceding month will be recruited from the community. Two recruitment strategies will be used: contacting men listed in primary care registers, and a community outreach method (time-space sampling). The intended sample of 798 men w...

This article conducts a comparative analysis of the interrelationship between climate, life expectancy and income between African and non-African countries. To put the analysis in a broader context of development, the paper develops an income-climate trap model that explains the multi-directional interaction between income, climate and life expectancy. It is suggested that the interaction can give rise to either a virtuous cycle of prosperity or a vicious cycle of poverty. Applying the model to a data set of 158 countries, we find that climate is a more important determinant of life expectancy in African countries than in non-African countries. We provide further empirical evidence that while climate is important in determining both life expectancy and income, income can in turn moderate the adverse effects of climate on life expectancy. In the past two decades, the income level of non-African countries has grown significantly while that of African countries has largely been stagnant, implying that the future development of African countries remains highly vulnerable to adverse climatic conditions. These findings have important implications in the context of climate change, as global warming is likely to create worsening climatic conditions that could see many less developed countries sinking deeper into an income-climate trap of underdevelopment in health.

ABSTRACT This study investigated the East Asia's automobile production network and its evolution from 1990 to 2010. A network analysis identified the major automobile trade links for parts and components (P&C) and final automobiles (FG), and changes in roles, trading partners and network patterns for the years 1990, 1995, 2000, 2005 and 2010. The results show that while the role of Malaysia and Singapore remained unchanged, China, Indonesia and Philippines expanded from importers to exporters of auto P&C. Thailand's role changed dramatically from an importer to an exporter of auto P&C and FG. Japan played a major role in the transformation of the industry during this period, as it started to import P&C from its East Asian partners and exported P&C to Thailand. Korea has followed in Japan's footsteps but with a focus on the industry's development in Vietnam and China.