STATISTICAL METHOD
Interrupted Time Series Analysis
Statistical Consulting for Intervention Impact
We apply interrupted time series analysis within your study design, fixing the intervention date and model in advance, allowing for autocorrelation and seasonality, and reporting level and slope changes with 95% confidence intervals.
No Commitment Required
Strictly Confidential
In-House PhD-Level Statisticians
Fixed Project Fee
Same-Day Response
METHOD OVERVIEW
What Is an Interrupted Time Series?
An interrupted time series evaluates a policy, program or event introduced at a known time by following one outcome over many periods before and after it. Segmented regression estimates the pre-intervention trend, then the change in level and the change in slope after the intervention, judged against the continuation of the earlier trend. The model must allow for autocorrelation and seasonality, and it needs enough time points on both sides. Without a control series, other events at the same time can be mistaken for the effect. OnlineSPSS, a specialist statistical consultancy for academic, clinical and professional research, provides interrupted time series analysis within your study design, with the relevant checks, effect estimates, uncertainty measures, interpretation and publication-ready reporting.
WHEN TO ENGAGE
When Is an Interrupted Time Series Appropriate?
The design fits an intervention at a known date with one outcome measured repeatedly, and these are the research situations in which researchers most often ask us to apply or confirm it.
Policy or Practice Change A guideline, formulary rule or regulation took effect on a known date, and routinely collected counts or rates exist for many periods around it.
Whole-Population Intervention Everyone was exposed at once, so no untreated group exists, and the trend before the change is the main benchmark for what would have followed.
Unplanned Event With a Known Date A drug shortage, a safety alert or a system outage disrupted practice without warning, and its effect on an outcome series is the question.
Reviewing a Completed ITS Model A segmented model has been fit and a reviewer has questioned autocorrelation or the number of points. Independent Statistical Review re-examines the model.
ANALYTICAL APPROACH
How We Apply Interrupted Time Series
Within the family of Clinical, Survival & Epidemiological, a time series model is specified around the intervention date and checked for the dependence that repeated measures bring.
Intervention Date and Form We fix the date, any lag or phase-in and whether a level change, a slope change or both are expected before post-intervention data are seen, so the result cannot choose the model.
Autocorrelation and Seasonality Ignoring autocorrelation usually understates uncertainty. We examine residual correlation, model seasonal patterns, and use autoregressive errors or Newey-West standard errors.
Series Length and Outcome Type Too few points before or after make a trend change hard to separate from noise. For counts we use Poisson or negative binomial models with an offset for the population at risk.
Control Series and Threats We look for a comparison series, such as an unaffected outcome or population, and list co-interventions and other events near the date that could produce the same pattern.
DELIVERABLES
What You Receive
As part of our Quantitative Data Analysis service, the series analysis is delivered with the estimates, diagnostics and files needed to report and defend it.
Data & Variable Review The measurement interval, the number of points either side, missing periods, changes in how the outcome is recorded and the intervention date are reviewed before modeling.
Method-Specific Checks Residual autocorrelation, seasonality, outliers and sensitivity to the intervention date and any lag are examined, with a control series analyzed where one exists.
Statistical Results The pre-intervention trend, the level change and slope change with 95% confidence intervals, and the estimated difference from the projected counterfactual at set follow-up times.
Publication-Ready Tables & Figures A time series plot with fitted segments and the projected counterfactual, and a results table, formatted to the journal's requirements, ready to insert.
Interpretation Changes are described as changes after the intervention date, with the strength of the case for attributing them to it stated against the known threats.
Reproducible Analysis Files R scripts, Stata do-files, SAS programs or SPSS syntax for the series model and diagnostics are provided where applicable, so every figure can be rerun if the data change.
ALTERNATIVE METHODS
When Another Method May Be More Appropriate
A single series with one interruption gives limited protection against other events. These methods fit when a comparison group exists, when the goal differs or when many sites share the change.
01
Difference-in-Differences (DiD)
When a comparison group was not exposed and both groups are measured around the change, comparing their changes guards against events that hit everyone.
02
Time Series Analysis
When the aim is forecasting, modeling dependence or decomposing a long series, and no intervention date is being evaluated, general time series models are the fit.
03
Linear Mixed Models (Multilevel Modeling)
When the same change is evaluated across many hospitals or regions, a multilevel model pools their series and lets the level and slope changes vary by site.
04
Regression Discontinuity (RD)
When treatment follows a cutoff on a score across many units, not a date in one series, comparing units on either side of the cutoff is the closer design.
CASE EXAMPLE
How We Work in Practice
An illustration of how a series analysis moves from a fixed intervention date to level and slope estimates and a planned comparison series.
ILLUSTRATIVE EXAMPLE · HOSPITAL PHARMACY OPERATIONS · UNITED STATES
A hospital pharmacy introduced an electronic prompt for a restricted antibiotic and wanted to evaluate whether monthly use changed after it went live.
The prompt applied to the whole hospital from one date, so no untreated group existed. Before post-launch data were reviewed, we fixed the date, a one-month phase-in and a model with a level change and a slope change. Monthly use per 1,000 patient-days was analyzed with a count model, seasonal terms and a check of residual autocorrelation. A stewardship program had also started near the launch, which we noted as a threat. We proposed a comparison series for an antibiotic the prompt did not cover, and the plan said how the two series would be compared.
Outcome: The pharmacy received the level and slope change estimates with 95% confidence intervals, the counterfactual figure, the model diagnostics, the comparison series results and the analysis code.
This example does not describe a real client or project. We never share details of actual client work, in line with the confidentiality we promise every researcher. This illustrative example is included to show the type, scope, and standard of work involved in this kind of analysis.
SOFTWARE
Statistical Software for Segmented Models
Segmented regression and time series diagnostics are run in the statistical software researchers already use, with syntax or scripts provided where applicable.
RESEARCH DISCIPLINES
Intervention Evaluation Across Research Fields
Interrupted time series designs appear across research fields, each with its own interventions, outcome series and reporting conventions.
A Defined Process for Your Analysis
We review your dataset, research question and design before any work begins, so the scope, deliverables, timeline and fixed project fee are clear from the outset.
01
Share Your Research Materials
Send your dataset, research question, study design and any existing analysis plan or output. We will let you know if anything else is needed to assess the project.
02
Receive Your Assessment
Your statistically trained Research Coordinator reviews the project and confirms the required analysis, deliverables, timeline and fixed project price before work begins.
03
Analysis & Delivery
Once you confirm the scope, an in-house statistician completes the agreed analysis and delivers the outputs, interpretation and reporting files.
RESEARCHER VOICE
What Researchers Say
Feedback from an epidemiologist who needed the assumptions and reporting choices worked through for a technical series problem.
“We came to the team with a fairly technical Interrupted Time Series problem: the repeated or paired structure of the measurements. They worked through the assumptions and reporting choices with us, then delivered a clean analytical summary rather than pages of software output. The final output was concise, clearly interpreted, and ready to use in our research reporting.”
RESEARCH RESPONSIBILITY
Statistical Support Without Taking Over Your Research
We provide the statistical work for your series analysis, and the study, its meaning and its authorship stay entirely with you.
Agreed Statistical Support We provide the agreed methodological, analytical and statistical reporting support for your intervention time series, within the scope confirmed before work begins.
No Intellectual Contribution OnlineSPSS claims no intellectual contribution to your research. Our work is statistical support, and authorship remains yours.
Your Research Decisions The research question, the choice of intervention date and your conclusions remain with you, and our interpretation stays within what the estimates and the design support.
Confidential Handling Research data and project details are handled confidentially and are not published. An NDA is available where required, and can be arranged before sensitive materials are shared.
FREQUENTLY ASKED QUESTIONS
Questions About Interrupted Time Series
Short answers to what researchers ask before evaluating an intervention over time, from series length to confidentiality and the assessment process.
How many time points do I need before and after?
No single number applies, because it depends on the noise, seasonality and the size of change you hope to detect. A common rule of thumb is at least 12 points on each side, and more when seasonality is present. With fewer points, trend changes are hard to separate from noise, so we discuss precision before modeling.
Should I use segmented regression or an ARIMA model?
Both can evaluate an interruption. Segmented regression estimates the level and slope change directly and handles autocorrelation with autoregressive terms or adjusted standard errors, which suits most policy and practice series. ARIMA models with intervention terms suit series with complex dependence or strong seasonality, but they need longer series and are harder to explain. We choose before modeling, from the length and pattern of the pre-intervention data.
Can this design show the intervention caused the change?
Not by itself. A single series is open to history threats, such as another policy or a change in recording near the date. A comparison series, a sharp timing of the change and a prespecified date strengthen the case, but the conclusion is worded as a change after the intervention, with the threats discussed.
Can you review an analysis that has already been completed?
Yes, we can review a completed series analysis before submission or a response to reviewers. We check the intervention date and lag, the model for level and slope change, the autocorrelation and seasonality handling, the count model, if used, and how the results were reported, and confirm what stands or needs to change.
How do you protect my dataset and research confidentiality?
Your dataset, output files and any protocol or analysis plan are handled confidentially, and project information is not published. An NDA is available where required and can be arranged before sensitive materials are shared. Read more about confidentiality and data security.
What happens when I request an assessment, and how long does it take?
You share your research question, design and relevant materials, and a statistically trained Research Coordinator reviews the request and responds the same day. The analysis, deliverables, timeline and one fixed project fee are then confirmed before any work begins, so the timeline follows from the agreed scope.
Still have a question? Find more answers in our full FAQ.
NEXT STEP
Need Support With an
Interrupted Time Series?
Share your research question, the intervention date and your series. We assess whether the design fits your evaluation and confirm the approach before any work begins.
No Commitment Required
Strictly Confidential
Fixed Project Fee
In-House PhD-Level Statisticians
Same-Day Response
Serving Researchers Since 2014