Loading presentation...

Present Remotely

Send the link below via email or IM


Present to your audience

Start remote presentation

  • Invited audience members will follow you as you navigate and present
  • People invited to a presentation do not need a Prezi account
  • This link expires 10 minutes after you close the presentation
  • A maximum of 30 users can follow your presentation
  • Learn more about this feature in our knowledge base article

Do you really want to delete this prezi?

Neither you, nor the coeditors you shared it with will be able to recover it again.


Interrupted Time Series Design

No description

Heather Halladay

on 12 February 2013

Comments (0)

Please log in to add your comment.

Report abuse

Transcript of Interrupted Time Series Design

Before and After . . . and After . . . and After . . . Interrupted Time Series Study Design Objectives What Is ITS? How Do We Use It? Breaking It Down:
Design Components Define Interrupted Time Series (ITS)
Discuss the rationale for its use
Review an ITS study
Examine an appraisal of an ITS study
Discuss ITS with classmates ITS involves the following:
Pre-intervention measurement (Ramsay et al., 2003)
The identification of a specific time point at which the intervention-- the interruption-- occurred (Ramsay et al., 2003)
There should be at least three data collection points (including pre-and post-intervention) (Gray, 2009).
Data collected at several time points following the intervention, allowing the researcher to assess
the effect of the intervention (Gray, 2009) www.ima.umn.edu Examples Studying the effect of a certain antibiotic
on methicilin-resistant staphyloccus aureus rates over a decade (Parienti, et al., 2011)
Examining the effects of policy or law introduction on morbidity or mortality (such as the effect of blood alcohol levels on alcohol-related fatal vehicle collisions) over several months or years Pause Please read the article attached to the Moodle discussion thread
before proceeding. You may went to stretch your legs a bit and grab
a coffee too! Quasi-experimental design
Effective design for
"investigating complex
interventions immediately
and over time" (Gray,
2009, p. 165)
Utilized when randomization is impractical
or impossible (Glass, Wilson, & Gottman,
2008; Ramsay, Matowe, Grilli,
Grimshaw & Thomas, 2003). The assumption is that observations post-intervention will be at a different slope or level than those occurring pre-intervention (Smith, 2009)
ITS should allow us to determine "whether the intervention has an effect significantly greater than th underlying secular trend" (Ramsay et al., 2003, p. 613)
Time-series examines causation: if a changes occurs in the series of observations after an intervention,
it can (potentially) be concluded that the change
was caused by the intervention
(Leedy & Ormrod, 2010). Challenges ITS studies "require some patience in data collection" (Glass et al., 2008, p. XIV)
Statistical analysis is complex (Glass et al., 2008)
The numerous observations recommended for data analysis can be difficult to obtain due to time constraints, missing or difficult-to-locate data (Smith, 2009)
The challenge of causation: it is always possible that another event occurred around the same time as the intervention. This unplanned event may have caused the change unknowingly, leading to erroneous
conclusions about causation (Leedy & Ormrod,
2010). ITS Appraisal More Design Components 1. Did the intervention occur independently of other changes? (questions adapted from Gray (2009) and Ramsay et al. (2003);
answers retrieved from Helder et al. 2012)) Yes. The authors indicated no other interventions to increase hand hygiene compliance were performed (p. 952). (Sources of bias to be explored in Seminar discussion.) 2. Was the intervention unlikely to affect
data collection? Yes. Methods of data collection were
the same both pre- and peri-
intervention. 3. Was the primary outcome assessed blindly or measured objectively? "The observers were not blinded to the outcome"
(p. 952) but objective measures of hand disinfection events were used (p. 952). 4. Was the primary outcome reliable or measured objectively? Yes. Two raters had agreement of kappa >0.8 (p. 952),
and see question 3. 5. Was the data set complete, covering 80-
100% of participants at each time point? Not specificied. 6. "Was the shape of the intervention effect specified?" (Gray, 2009, p. 167) The authors theorized about the change in com-pliance, which would determine the shape (p. 953). 7. Was the rationale for number and spacing of data points described? Yes. The weekly spacing was described, with rationale
being to "determine the longitudinal effects and avoid
autocorrelation" (Helder et al., 2012, p. 953). 8. Was the study analyzed using
techniques appropriate for time-series? Yes, "regression analysis of interrupted time
series data" was used (Helder et al.,
2012, p. 953). So . . . Is This A
High Quality Study? The study meets most, though not
all of the criteria set out by Gray (2009)
and Hedler et al. (2012). There are
some design weaknesses and limitations. The next mountain to climb . . . return to the Moodle discussion thread to dialogue
about the quality of the study and other
aspects of ITS. References Glass, G., Wilson, V., & Gottman, J. (2008). Design and analysis of time-series
experiments [Preview]. Retrieved from http://books.google.ca
Gray, M. (2009). Evidenced-based healthcare and public health: How to make decisions
about health services and public health (3rd ed.). New York, NY: Elsevier.
Helder, O., Weggelaar, Q., Waarsenburg, D., Looman, C., van Goudoever, J., Brug, J., &
Kornelisse, R. (2012). Computer screen saver hand hygiene information curbs a negative trend in hand hygiene behavior. American Journal of Infection Control, 40(10), 951-954. doi:10.1016/j.ajic.2011.12.003
Leedy, P., & Ormrod, J. (2010). Practical research: Planning and design (9th ed.). Upper
Saddle River, NJ: Pearson.
Parienti, J., Cattoir, V., Thibon, P., Lebouvier, G., Verdon, R., Daubin, C., . . .
Charbonneau, P. (2011). Hospital-wide modification of fluoroquinolone policy and meticilin-resistant Staphylococcus aureus rates: A 10-year interrupted time-series analysis. Journal of Hospital Infection, 78(4), 118-122. Retrieved from www.sciencedirect.com
Ramsay, C., Matowe, L., Grilli, R., Grimshaw, J., & Thomas, R. (2003). Interrupted time
series designs in health technology assessment: Lessons from two systematic reviews of behavior change strategies. International Journal of Technology Assessment in Health Care, 19(4), 613-623. doi:10.1017/S0266462303000576
Smith, K. (2009). Interrupted time series: What, why and how [Power Point
presentation]. Retrieved from
http://www.scribd.com/doc/21794785/Interrupted-Time-Series Thanks for your time and attention. Heather Halladay
MHST 610
February 12, 2013
Full transcript