Interrupted time series analysis is increasingly used to evaluate the impact of large-scale health interventions. While segmented regression is a common approach, it is not always adequate, ...especially in the presence of seasonality and autocorrelation. An Autoregressive Integrated Moving Average (ARIMA) model is an alternative method that can accommodate these issues.
We describe the underlying theory behind ARIMA models and how they can be used to evaluate population-level interventions, such as the introduction of health policies. We discuss how to select the shape of the impact, the model selection process, transfer functions, checking model fit, and interpretation of findings. We also provide R and SAS code to replicate our results.
We illustrate ARIMA modelling using the example of a policy intervention to reduce inappropriate prescribing. In January 2014, the Australian government eliminated prescription refills for the 25 mg tablet strength of quetiapine, an antipsychotic, to deter its prescribing for non-approved indications. We examine the impact of this policy intervention on dispensing of quetiapine using dispensing claims data.
ARIMA modelling is a useful tool to evaluate the impact of large-scale interventions when other approaches are not suitable, as it can account for underlying trends, autocorrelation and seasonality and allows for flexible modelling of different types of impacts.
To present updated national birthweight percentiles by gestational age for male and female singleton infants born in Australia.
Cross-sectional population-based study of 2.53 million singleton live ...births in Australia between 1998 and 2007.
Birthweight percentiles by gestational age and sex.
Between 1998 and 2007, women in Australia gave birth to 2 539 237 live singleton infants. Of these, 2 537 627 had a gestational age between 20 and 44 weeks, and sex and birthweight data were available. Birthweight percentiles are presented by sex and gestational age for a total of 2 528 641 births, after excluding 8986 infants with outlying birthweights. Since the publication of the previous Australian birthweight percentiles in 1999, median birthweight for term babies has increased between 0 and 25 g for boys and between 5 g and 45 g for girls.
There has been only a small increase in birthweight percentiles for babies of both sexes and most gestational ages since 1991-1994. These national percentiles provide a current Australian reference for clinicians and researchers assessing weight at birth.
Abstract
Background
Obtaining unbiased causal estimates from longitudinal observational data can be difficult due to exposure-affected time-varying confounding. The past decade has seen considerable ...development in methods for analysing such complex longitudinal data. However, the extent to which those methods have been implemented is unclear. This study describes and characterizes the state of the field in methods adjusting for exposure-affected time-varying confounding, and examines their use in the literature.
Methods
We systematically reviewed the literature from 2000 to 2016 for studies adjusting for time-dependent confounding, including use of specific methods like inverse probability of treatment weighting (IPTW). Articles were coded based on the methods used and, for applied articles, the topic areas covered.
Results
We screened 4239 abstracts, and subsequently reviewed 1100 articles, leaving 542 relevant articles in the analyses. The number of published articles increased from two in 2000, to 112 in 2016. This increase was primarily in applied articles using IPTW, which increased from one study in 2000, to 90 in 2016. Of the 432 studies with applications to observed data, 60.9% were on at least one of: HIV (30.6%), cardiopulmonary health (13.2%), kidney disease (11.8%) or mental health (10.0%).
Conclusions
There has been marked growth in reports addressing exposure-affected time-varying confounding. This was driven by work in a small number of topic areas, with other areas showing relatively little uptake. In addition, despite developments in more advanced methods such doubly robust techniques and estimation via machine learning, implementation has been largely concentrated on the simpler, yet potentially less robust, IPTW.
To investigate the effectiveness of a centralized, nurse-delivered telephone-based service to improve care coordination and patient-reported outcomes after surgery for colorectal cancer.
Patients ...with a newly diagnosed colorectal cancer were randomly assigned to the CONNECT intervention or usual care. Intervention-group patients received standardized calls from the centrally based nurse 3 and 10 days and 1, 3, and 6 months after discharge from hospital. Unmet supportive care needs, experience of care coordination, unplanned readmissions, emergency department presentations, distress, and quality of life (QOL) were assessed by questionnaire at 1, 3, and 6 months.
Of 775 patients treated at 23 public and private hospitals in Australia, 387 were randomly assigned to the intervention group and 369 to the control group. There were no significant differences between groups in unmet supportive care needs, but these were consistently low in both groups at both follow-up time points. There were no differences between the groups in emergency department presentations (10.8% v 13.8%; P = .2) or unplanned hospital readmissions (8.6% v 10.5%; P = .4) at 1 month. By 6 months, 25.6% of intervention-group patients had reported an unplanned readmission compared with 27.9% of controls (P = .5). There were no significant differences in experience of care coordination, distress, or QOL between groups at any follow-up time point.
This trial failed to demonstrate substantial benefit of a centralized system to provide standardized, telephone follow-up for postoperative patients with colorectal cancer. Future interventions could investigate a more tailored approach.
To investigate whether the introduction of tobacco plain packaging in Australia from 1 October 2012 was associated with a change in the number of calls to the smoking cessation helpline, Quitline, ...and to compare this with the impact of the introduction of graphic health warnings from 1 March 2006.
Whole-of-population interrupted time-series analysis in New South Wales and the Australian Capital Territory between 1 March 2005 and October 2006 for the comparator, graphic health warnings, and October 2011 and April 2013 for the intervention of interest, tobacco plain packaging.
Weekly number of calls to the Quitline, after adjusting for seasonal trends, anti-tobacco advertising, cigarette costliness and the number of smokers in the community.
There was a 78% increase in the number of calls to the Quitline associated with the introduction of plain packaging (baseline, 363/week; peak, 651/week 95% CI, 523-780/week; P < 0.001). This peak occurred 4 weeks after the initial appearance of plain packaging and has been prolonged. The 2006 introduction of graphic health warnings had the same relative increase in calls (84%; baseline, 910/week; peak, 1673/week 95% CI, 1383-1963/week; P < 0.001) but the impact of plain packaging has continued for longer.
There has been a sustained increase in calls to the Quitline after the introduction of tobacco plain packaging. This increase is not attributable to anti-tobacco advertising activity, cigarette price increases nor other identifiable causes. This is an important incremental step in comprehensive tobacco control.
ABSTRACT BACKGROUND Australia introduced tamper-resistant controlled-release (CR) oxycodone in April 2014. We quantified the impact of the reformulation on dispensing, switching and poisonings. ...METHODS We performed interrupted time-series analyses using population-representative national dispensing data from 2012 to 2016. We measured dispensing of oxycodone CR (≥ 10 mg), discontinuation of use of strong opioids and switching to other strong opioids after the reformulation compared with a historical control period. Similarly, we compared calls about intentional opioid poisoning using data from a regional poisons information centre. RESULTS After the reformulation, dispensing decreased for 10–30 mg (total level shift −11.1%, 95% confidence interval CI, −17.2% to −4.6%) and 40–80 mg oxycodone CR (total level shift −31.5%, 95% CI −37.5% to −24.9%) in participants less than 65 years of age but was unchanged in people 65 years of age or older. Compared with the previous year, discontinuation of use of strong opioids did not increase (adjusted hazard ratio HR 0.95, 95% CI 0.91 to 1.00), but switching to oxycodone/naloxone did increase (adjusted HR 1.54, 95% CI 1.32 to 1.79). Switching to morphine varied by age ( p < 0.001), and the greatest increase was in participants less than 45 years of age (adjusted HR 4.33, 95% CI 2.13 to 8.80). Participants switching after the reformulation were more likely to be dispensed a tablet strength of 40 mg or more (adjusted odds ratio OR 1.40, 95% CI 1.09 to 1.79). Calls for intentional poisoning that involved oxycodone taken orally increased immediately after the reformulation (incidence rate ratio (IRR) 1.31, 95% CI 1.05–1.64), but there was no change for injected oxycodone. INTERPRETATION The reformulation had a greater impact on opioid access patterns of people less than 65 years of age who were using higher strengths of oxycodone CR. This group has been identified as having an increased risk of problematic opioid use and warrants closer monitoring in clinical practice.
Routinely collected data have been increasingly used to assess long‐term opioid therapy (LTOT) patterns, with very little guidance on how to measure LTOT from these data sources. We conducted a ...systematic review of studies published between January 2000 and July 2019 to catalogue LTOT definitions, the rationale for definitions and LTOT rates in observational research using routinely collected data in nonsurgical settings. We screened 4056 s, 210 full‐text manuscripts and included 128 studies, mostly from the United States (81%) and published between 2015 and 2019 (69%). We identified 78 definitions of LTOT, commonly operationalised as 90 days of use within a year (23%). Studies often used multiple criteria to derive definitions (60%), mostly based on measures of duration, such as supply days/days of use (66%), episode length (21%) or prescription fills within specified time periods (12%). Definitions were based on previous publications (63%), clinical judgment (16%) or empirical data (3%); 10% of studies applied more than one definition. LTOT definition was not provided with enough details for replication in 14 studies and 38 studies did not specify the opioids evaluated. Rates of LTOT within study populations ranged from 0.2% to 57% according to study design and definition used. We observed a substantial rise in the last 5 years in studies evaluating LTOT with large variability in the definitions used and poor reporting of the rationale and implementation of definitions. This variation impacts on research reproducibility, comparability of findings and the development of strategies aiming to curb therapy that is not guideline‐recommended.
Background and Aims
Unhealthy alcohol consumption is a key concern for Aboriginal and Torres Strait Islander (‘Indigenous’) communities. It is important to identify and treat at‐risk drinkers, to ...prevent harms to physical or social wellbeing. We aimed to test whether training and support for Aboriginal Community Controlled Health Service (ACCHS) staff would increase rates of alcohol screening and brief intervention.
Design
Cluster randomized trial.
Setting
Australia.
Cases/Intervention/Measurements
Twenty‐two ACCHSs that see at least 1000 clients per year and use Communicare as practice management software. The study included data on 70 419 clients, training, regular data feedback, collaborative support and funding for resources ($9000). Blinding was not used. The comparator was waiting‐list control (equal allocation). Alcohol Use Disorder Identification Test (AUDIT‐C) screening and records of brief interventions were extracted from practice management software at 2‐monthly intervals. Observations described the clinical actions taken for clients over each 2‐month interval. The baseline period (28 August 2016–28 August 2017) was compared with the post‐implementation period (29 August 2017–28 August 2018). We used multi‐level logistic regression to test the hypotheses that clients attending a service receiving active support would be more likely to be screened with AUDIT‐C (primary outcome) or to receive a brief intervention (secondary outcome).
Findings
We observed an increase in the odds of screening with AUDIT‐C for both groups, but the increase was 5.52 95% confidence interval (CI) = 4.31, 7.07 times larger at services receiving support. We found little evidence that the support programme increased the odds of a recorded brief intervention relative to control services (odds ratio = 2.06; 95% CI = 0.90, 4.69). Differences in baseline screening activity between treatment and control reduce the certainty of our findings.
Conclusions
Providing Aboriginal Community Controlled Health Services with training and support can improve alcohol (AUDIT‐C) screening rates.
Objective:This study compared changes in criminal justice contact, quality of life, and social connectedness over a 12-month follow-up period between participants in two Housing First configurations ...(scattered site SS and congregate site CS).Methods:A longitudinal, quantitative design was utilized for this ecological study. Changes in individual outcomes over time were compared for SS and CS participants who completed both baseline and 12-month follow-up surveys (N=63).Results:The number of contacts with various types of criminal justice system channels differed significantly between SS and CS participants, decreasing significantly among SS participants and increasing significantly among CS participants. The two groups did not differ on quality-of-life outcomes or social-connectedness measures, with the exception of case management engagement, whereby a greater proportion of SS participants disengaged from this service over time compared with CS participants. At follow-up, significant within-group changes over time emerged, with increased boredom reported among SS participants, whereas CS participants reported improvements in social relationships, with fewer reporting losing their temper.Conclusions:The findings supported the notion that the Housing First approach has the potential to significantly improve the lives of persons who have experienced chronic homelessness, a traditionally marginalized and vulnerable group. Over time, this may result in a reduction in the use of acute services, thereby reducing societal costs. The challenge remains to identify the suitability of particular configurations of housing and support and how service delivery can optimize individual outcomes so positive outcomes are maintained in the longer term.
Abstract This study provides Australian data on the incidence of human papillomavirus (HPV)-related oropharyngeal cancer to aid the debate on extending the HPV vaccination programme to males. The HPV ...status for 302 oropharyngeal cancers diagnosed between 1987 and 2006 was determined by HPV E6-targeted multiplex real-time PCR/p16 immunohistochemistry. The overall HPV-positivity rate was 36% (94% types 16 and 18). HPV-related cancer increased from 19% (1987–1990) to 47% (2001–2005). HPV data used in conjunction with Australian cancer incidence data 2001–2005 showed that 1.56 cases of oropharyngeal cancer per 100,000 males per year were associated with HPV types targeted by the vaccine. Vaccinating males may substantially reduce the burden of oropharyngeal cancer in Australia.
PDF
