scholarly journals Comparison of six statistical methods for interrupted time series studies: empirical evaluation of 190 published series

2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Simon L. Turner ◽  
Amalia Karahalios ◽  
Andrew B. Forbes ◽  
Monica Taljaard ◽  
Jeremy M. Grimshaw ◽  
...  
Keyword(s):  
Time Series ◽  
Statistical Method ◽  
Statistical Methods ◽  
Time Series Data ◽  
Statistical Significance ◽  
Empirical Evaluation ◽  
Interrupted Time Series ◽  
Series Data ◽  
Standard Errors ◽  
The Impact

Abstract Background The Interrupted Time Series (ITS) is a quasi-experimental design commonly used in public health to evaluate the impact of interventions or exposures. Multiple statistical methods are available to analyse data from ITS studies, but no empirical investigation has examined how the different methods compare when applied to real-world datasets. Methods A random sample of 200 ITS studies identified in a previous methods review were included. Time series data from each of these studies was sought. Each dataset was re-analysed using six statistical methods. Point and confidence interval estimates for level and slope changes, standard errors, p-values and estimates of autocorrelation were compared between methods. Results From the 200 ITS studies, including 230 time series, 190 datasets were obtained. We found that the choice of statistical method can importantly affect the level and slope change point estimates, their standard errors, width of confidence intervals and p-values. Statistical significance (categorised at the 5% level) often differed across the pairwise comparisons of methods, ranging from 4 to 25% disagreement. Estimates of autocorrelation differed depending on the method used and the length of the series. Conclusions The choice of statistical method in ITS studies can lead to substantially different conclusions about the impact of the interruption. Pre-specification of the statistical method is encouraged, and naive conclusions based on statistical significance should be avoided.

scholarly journals Comparison of Six Statistical Methods for Interrupted time Series Studies: Empirical Evaluation of 190 Published Series

2020 ◽  
Author(s):  
Simon Turner ◽  
Amalia Karahalios ◽  
Andrew Forbes ◽  
Monica Taljaard ◽  
Jeremy Grimshaw ◽  
...  
Keyword(s):  
Time Series ◽  
Statistical Method ◽  
Statistical Methods ◽  
Statistical Significance ◽  
Empirical Evaluation ◽  
Interrupted Time Series ◽  
Series Data ◽  
Standard Errors ◽  
P Values ◽  
The Impact

Abstract Background The Interrupted Time Series (ITS) is a quasi-experimental design commonly used in public health to evaluate the impact of interventions or exposures. Multiple statistical methods are available to analyse data from ITS studies, but no empirical investigation has examined how the different methods compare when applied to real-world datasets. MethodsA random sample of 200 ITS studies identified in a previous methods review were included. Time series data from each of these studies was sought. Each dataset was re-analysed using six statistical methods. Point and confidence interval estimates for level and slope changes, standard errors, p-values and estimates of autocorrelation were compared between methods. ResultsFrom the 200 ITS studies, including 230 time series, 190 datasets were obtained. We found that the choice of statistical method can importantly affect the level and slope change point estimates, their standard errors, width of confidence intervals and p-values. Statistical significance (categorised at the 5% level) often differed across the pairwise comparisons of methods, ranging from 4% to 25% disagreement. Estimates of autocorrelation differed depending on the method used and the length of the series. ConclusionsThe choice of statistical method in ITS studies can lead to substantially different conclusions about the impact of the interruption. Pre-specification of the statistical method is encouraged, and naive conclusions based on statistical significance should be avoided.

scholarly journals Design characteristics and statistical methods used in interrupted time series studies evaluating public health interventions: protocol for a review

BMJ Open ◽  
2019 ◽  
Vol 9 (1) ◽  
pp. e024096 ◽  
Author(s):  
Simon L Turner ◽  
Amalia Karahalios ◽  
Andrew B Forbes ◽  
Monica Taljaard ◽  
Jeremy M Grimshaw ◽  
...  
Keyword(s):  
Public Health ◽  
Time Series ◽  
Statistical Methods ◽  
Time Series Data ◽  
Interrupted Time Series ◽  
Health Interventions ◽  
Series Data ◽  
Public Health Interventions ◽  
Design Characteristics ◽  
The Impact

IntroductionAn interrupted time series (ITS) design is an important observational design used to examine the effects of an intervention or exposure. This design has particular utility in public health where it may be impracticable or infeasible to use a randomised trial to evaluate health system-wide policies, or examine the impact of exposures (such as earthquakes). There have been relatively few studies examining the design characteristics and statistical methods used to analyse ITS designs. Further, there is a lack of guidance to inform the design and analysis of ITS studies.This is the first study in a larger project that aims to provide tools and guidance for researchers in the design and analysis of ITS studies. The objectives of this study are to (1) examine and report the design characteristics and statistical methods used in a random sample of contemporary ITS studies examining public health interventions or exposures that impact on health-related outcomes, and (2) create a repository of time series data extracted from ITS studies. Results from this study will inform the remainder of the project which will investigate the performance of a range of commonly used statistical methods, and create a repository of input parameters required for sample size calculation.Methods and analysisWe will collate 200 ITS studies evaluating public health interventions or the impact of exposures. ITS studies will be identified from a search of the bibliometric database PubMed between the years 2013 and 2017, combined with stratified random sampling. From eligible studies, we will extract study characteristics, details of the statistical models and estimation methods, effect metrics and parameter estimates. Further, we will extract the time series data when available. We will use systematic review methods in the screening, application of inclusion and exclusion criteria, and extraction of data. Descriptive statistics will be used to summarise the data.Ethics and disseminationEthics approval is not required since information will only be extracted from published studies. Dissemination of the results will be through peer-reviewed publications and presentations at conferences. A repository of data extracted from the published ITS studies will be made publicly available.

Abstract 19225: Impact of Change in Resuscitation Guidelines on National Out-of-hospital Cardiac Arrest Outcomes: Fulfilled Expectations?

Circulation ◽  
2015 ◽  
Vol 132 (suppl_3) ◽  
Author(s):  
Shaker M Eid ◽  
Aiham Albaeni ◽  
Rebeca Rios ◽  
May Baydoun ◽  
Bolanle Akinyele ◽  
...  
Keyword(s):  
Time Series ◽  
Cardiac Arrest ◽  
Time Series Data ◽  
Statistical Significance ◽  
Interrupted Time Series ◽  
The United States ◽  
Series Data ◽  
National Database ◽  
Resuscitation Guidelines ◽  
Hospital Cardiac Arrest

Background: The intent of the 5-yearly Resuscitation Guidelines is to improve outcomes. Previous studies have yielded conflicting reports of a beneficial impact of the 2005 guidelines on out-of-hospital cardiac arrest (OHCA) survival. Using a national database, we examined survival before and after the introduction of both the 2005 and 2010 guidelines. Methods: We used the 2000 through 2012 National Inpatient Sample database to select patients ≥18 years admitted to hospitals in the United States with non-traumatic OHCA (ICD-9 CM codes 427.5 & 427.41). A quasi-experimental (interrupted time series) design was used to compare monthly survival trends. Outcomes for OHCA were compared pre- and post- 2005 and 2010 resuscitation guidelines release as follows: 01/2000-09/2005 vs. 10/2005-9/2010 and 10/2005-9/2010 vs. 10/2010-12/2012. Segmented regression analyses of interrupted time series data were performed to examine changes in survival to hospital discharge. Results: For the pre- and post- guidelines periods, 81600, 69139 and 36556 patients respectively survived to hospital admission following OHCA. Subsequent to the release of the 2005 guidelines, there was a statistically significant worsening in survival trends (β= -0.089, 95% CI -0.163 – -0.016, p =0.018) until the release of the 2010 guidelines when a sharp increase in survival was noted which persisted for the period of study (β= 0.054, 95% CI -0.143 – 0.251, p =0.588) but did not achieve statistical significance (Figure). Conclusion: National clinical guidelines developed to impact outcomes must include mechanisms to assess whether benefit actually occurs. The worsening in OHCA survival following the 2005 guidelines is thought provoking but the improvement following the release of the 2010 guidelines is reassuring and worthy of perpetuation.

scholarly journals Massive social protests amid the pandemic in selected Colombian cities: Did they increase COVID-19 cases?

2021 ◽  
Author(s):  
Jose Moreno-Montoya ◽  
Laura A Rodriguez Villamizar ◽  
Alvaro Javier Idrovo
Keyword(s):  
Time Series ◽  
Time Series Data ◽  
Moving Average ◽  
Interrupted Time Series ◽  
Series Data ◽  
Arima Models ◽  
Interrupted Time Series Analysis ◽  
Social Protests ◽  
Show Evidence ◽  
The Impact

Background. Since April 28, 2021, in Colombia there are social protests with numerous demonstrations in various cities. This occurs whereas the country faces the third wave of the COVID-19 pandemic. The aim of this study was to assess the effect of social protests on the number and trend of the confirmed COVID-19 cases in some selected Colombian cities where social protests had more intensity. Methods. We performed and interrupted time-series analysis (ITSA) and Autoregressive Integrated Moving Average (ARIMA) models, based on the confirmed COVID-19 cases in Colombia, between March 1 and May 15, 2021, for the cities of Bogota, Cali, Barranquilla, Medellin, and Bucaramanga. The ITSA models estimated the impact of social demonstrations on the number and trend of cases for each city by using Newey-West standard errors and ARIMA models assessed the overall pattern of the series and effect of the intervention. We considered May 2, 2021, as the intervention date for the analysis, five days after social demonstrations started in the country. Findings. During the study period the number of cases by city was 1,014,815 for Bogota, 192,320 for Cali, 175,269 for Barranquilla, 311,904 for Medellin, and 62,512 for Bucaramanga. Heterogeneous results were found among cities. Only for the cities of Cali and Barranquilla statistically significant changes in trend of the number of cases were obtained after the intervention: positive in the first city, negative in the second one. None ARIMA models show evidence of abrupt changes in the trend of the series for any city and intervention effect was only positive for Bucaramanga. Interpretation. The findings confer solid evidence that social protests had an heterogenous effect on the number and trend of COVID-19 cases. Divergent effects might be related to the epidemiologic time of the pandemic and the characteristics of the social protests. Assessing the effect of social protests within a pandemic is complex and there are several methodological limitations. Further analyses are required with longer time-series data.

EFFECTS OF TREND AND PERIODICITY ON THE CORRELATION DIMENSION AND THE LYAPUNOV EXPONENTS

2008 ◽  
Vol 18 (12) ◽  
pp. 3679-3687 ◽  
Author(s):  
AYDIN A. CECEN ◽  
CAHIT ERKAL
Keyword(s):  
Time Series ◽  
Lyapunov Exponent ◽  
Lyapunov Exponents ◽  
Correlation Dimension ◽  
Time Series Data ◽  
Logistic Map ◽  
Model Identification ◽  
Series Data ◽  
Largest Lyapunov Exponent ◽  
The Impact

We present a critical remark on the pitfalls of calculating the correlation dimension and the largest Lyapunov exponent from time series data when trend and periodicity exist. We consider a special case where a time series Zi can be expressed as the sum of two subsystems so that Zi = Xi + Yi and at least one of the subsystems is deterministic. We show that if the trend and periodicity are not properly removed, correlation dimension and Lyapunov exponent estimations yield misleading results, which can severely compromise the results of diagnostic tests and model identification. We also establish an analytic relationship between the largest Lyapunov exponents of the subsystems and that of the whole system. In addition, the impact of a periodic parameter perturbation on the Lyapunov exponent for the logistic map and the Lorenz system is discussed.

scholarly journals A Comparison of Time-Series Predictions for Healthcare Emergency Department Indicators and the Impact of COVID-19

2021 ◽  
Vol 11 (8) ◽  
pp. 3561
Author(s):  
Diego Duarte ◽  
Chris Walshaw ◽  
Nadarajah Ramesh
Keyword(s):  
Time Series ◽  
Time Series Data ◽  
Moving Average ◽  
General Regression Neural Network ◽  
Series Data ◽  
Machine Learning Model ◽  
Hospital Patients ◽  
Major Incidents ◽  
General Regression ◽  
The Impact

Across the world, healthcare systems are under stress and this has been hugely exacerbated by the COVID pandemic. Key Performance Indicators (KPIs), usually in the form of time-series data, are used to help manage that stress. Making reliable predictions of these indicators, particularly for emergency departments (ED), can facilitate acute unit planning, enhance quality of care and optimise resources. This motivates models that can forecast relevant KPIs and this paper addresses that need by comparing the Autoregressive Integrated Moving Average (ARIMA) method, a purely statistical model, to Prophet, a decomposable forecasting model based on trend, seasonality and holidays variables, and to the General Regression Neural Network (GRNN), a machine learning model. The dataset analysed is formed of four hourly valued indicators from a UK hospital: Patients in Department; Number of Attendances; Unallocated Patients with a DTA (Decision to Admit); Medically Fit for Discharge. Typically, the data exhibit regular patterns and seasonal trends and can be impacted by external factors such as the weather or major incidents. The COVID pandemic is an extreme instance of the latter and the behaviour of sample data changed dramatically. The capacity to quickly adapt to these changes is crucial and is a factor that shows better results for GRNN in both accuracy and reliability.

scholarly journals Assessing Stream-Aquifer Connectivity in a Coastal California Watershed

Water ◽  
2021 ◽  
Vol 13 (4) ◽  
pp. 416
Author(s):  
Bwalya Malama ◽  
Devin Pritchard-Peterson ◽  
John J. Jasbinsek ◽  
Christopher Surfleet
Keyword(s):  
Time Series ◽  
Spectral Analysis ◽  
Water Level ◽  
Time Series Data ◽  
Series Data ◽  
Low Permeability ◽  
Pumping Tests ◽  
Stream Flows ◽  
The Impact

We report the results of field and laboratory investigations of stream-aquifer interactions in a watershed along the California coast to assess the impact of groundwater pumping for irrigation on stream flows. The methods used include subsurface sediment sampling using direct-push drilling, laboratory permeability and particle size analyses of sediment, piezometer installation and instrumentation, stream discharge and stage monitoring, pumping tests for aquifer characterization, resistivity surveys, and long-term passive monitoring of stream stage and groundwater levels. Spectral analysis of long-term water level data was used to assess correlation between stream and groundwater level time series data. The investigations revealed the presence of a thin low permeability silt-clay aquitard unit between the main aquifer and the stream. This suggested a three layer conceptual model of the subsurface comprising unconfined and confined aquifers separated by an aquitard layer. This was broadly confirmed by resistivity surveys and pumping tests, the latter of which indicated the occurrence of leakage across the aquitard. The aquitard was determined to be 2–3 orders of magnitude less permeable than the aquifer, which is indicative of weak stream-aquifer connectivity and was confirmed by spectral analysis of stream-aquifer water level time series. The results illustrate the importance of site-specific investigations and suggest that even in systems where the stream is not in direct hydraulic contact with the producing aquifer, long-term stream depletion can occur due to leakage across low permeability units. This has implications for management of stream flows, groundwater abstraction, and water resources management during prolonged periods of drought.

Welfare Use in Japan: Trends and Determinants

2007 ◽  
pp. 88
Author(s):  
Wataru Suzuki ◽  
Yanfei Zhou
Keyword(s):  
Time Series ◽  
Longitudinal Data ◽  
Time Series Data ◽  
Positive Impact ◽  
Series Data ◽  
Negative Effects ◽  
Long Time ◽  
Welfare Use ◽  
The Impact

This article represents the first step in filling a large gap in knowledge concerning why Public Assistance (PA) use recently rose so fast in Japan. Specifically, we try to address this problem not only by performing a Blanchard and Quah decomposition on long-term monthly time series data (1960:04-2006:10), but also by estimating prefecturelevel longitudinal data. Two interesting findings emerge from the time series analysis. The first is that permanent shock imposes a continuously positive impact on the PA rate and is the main driving factor behind the recent increase in welfare use. The second finding is that the impact of temporary shock will last for a long time. The rate of the use of welfare is quite rigid because even if the PA rate rises due to temporary shocks, it takes about 8 or 9 years for it to regain its normal level. On the other hand, estimations of prefecture-level longitudinal data indicate that the Financial Capability Index (FCI) of the local government2 and minimum wage both impose negative effects on the PA rate. We also find that the rapid aging of Japan's population presents a permanent shock in practice, which makes it the most prominent contribution to surging welfare use.

The Impact of a National Antimicrobial Stewardship Program on Antibiotic Prescribing in Primary Care: An Interrupted Time Series Analysis

2018 ◽  
Vol 69 (2) ◽  
pp. 227-232 ◽  
Author(s):  
Violeta Balinskaite ◽  
Alan P Johnson ◽  
Alison Holmes ◽  
Paul Aylin
Keyword(s):  
Primary Care ◽  
Time Series ◽  
Broad Spectrum ◽  
Interrupted Time Series ◽  
Series Data ◽  
Antibiotic Prescribing ◽  
Segmented Regression ◽  
Broad Spectrum Antibiotic ◽  
Stewardship Program ◽  
The Impact

Abstract Background The Quality Premium was introduced in 2015 to financially reward local commissioners of healthcare in England for targeted reductions in antibiotic prescribing in primary care. Methods We used a national antibiotic prescribing dataset from April 2013 until February 2017 to examine the number of antibiotic items prescribed, the total number of antibiotic items prescribed per STAR-PU (specific therapeutic group age/sex-related prescribing units), the number of broad-spectrum antibiotic items prescribed, and broad-spectrum antibiotic items prescribed, expressed as a percentage of the total number of antibiotic items. To evaluate the impact of the Quality Premium on antibiotic prescribing, we used a segmented regression analysis of interrupted time series data. Results During the study period, over 140 million antibiotic items were prescribed in primary care. Following the introduction of the Quality Premium, antibiotic items prescribed decreased by 8.2%, representing 5933563 fewer antibiotic items prescribed during the 23 post-intervention months, as compared with the expected numbers based on the trend in the pre-intervention period. After adjusting for the age and sex distribution in the population, the segmented regression model also showed a significant relative decrease in antibiotic items prescribed per STAR-PU. A similar effect was found for broad-spectrum antibiotics (comprising 10.1% of total antibiotic prescribing), with an 18.9% reduction in prescribing. Conclusions This study shows that the introduction of financial incentives for local commissioners of healthcare to improve the quality of prescribing was associated with a significant reduction in both total and broad-spectrum antibiotic prescribing in primary care in England.

scholarly journals Political Stability and its Impact on Economic Growth of Pakistan (1988-2018): A Time Series Analysis

2020 ◽  
Vol 6 (1) ◽  
pp. 273-282
Author(s):  
Majid Hussain Phul ◽  
Muhammad Saleem Rahpoto ◽  
Ghulam Muhammad Mangnejo
Keyword(s):  
Economic Growth ◽  
Time Series ◽  
Time Series Data ◽  
Political Stability ◽  
Series Data ◽  
Explanatory Variables ◽  
Long Run ◽  
Short Run ◽  
The Impact ◽  
Positive Effect

This research paper empirically investigates the outcome of Political stability on economic growth (EG) of Pakistan for the period of 1988 to 2018. Political stability (PS), gross fixed capital formation (GFCF), total labor force (TLF) and Inflation (INF) are important explanatory variables. Whereas for model selection GDPr is used as the dependent variable. To check the stationary of time series data Augmented Dickey Fuller (ADF) unit root (UR) test has been used,  and whereas to find out the long run relationship among variables, OLS method has been used. The analysis the impact of PS on EG (EG) in the short run, VAR model has been used. The outcomes show that all the variables (PS, GFCF, TLF and INF) have a significantly positive effect on the EG of Pakistan in the long run period. But the effect of PS on GDP is smaller. Further, in this research we are trying to see the short run relationship between GDP and other explanatory variables. The outcomes show that PS does not have such effect on GDP in the short run analysis. While GFCF, TLF and INF have significantly positive effect on GDP of Pakistan in the short run period.

Sign in / Sign up

Export Citation Format

Share Document