Tag Archives: quasi-experimental

“Should you? Can you?”

ApplesOranges2Quasi-experiments are a lot of work, yet don’t have the same scientific power to show cause and effect, as do randomized controlled trials (RCTs).   An RCT would provide better support for any hypothesis that X causes Y.   [As a quick review of what quasi-experimental versus RCT studies are, see “Of Mice & Cheese” and/or “Out of Control (Groups).”]

So why do quasi-experimental studies at all?  Why not always do RCTs when we are testing cause and effect?  Here are 3 reasons:

#1  Sometimes ETHICALLY the researcher canNOT randomly assign subjects to a control Smokingand an experimental group.  If the researcher wants to compare health outcomes of smokers with non-smokers, the researcher cannot assign some people to smoke and others not to smoke!  Why?  Because we already know that smoking has significant harmful effects. (Of course, in a dictatorship, by using the police a researcher could assign them to smoke or not smoke, but I don’t think we wanna go there.)

#2 Sometimes PHYSICALLY the researcher canNOT randomly assign subjects to control & Country of Originexperimental groups.   If the researcher wants to compare health outcomes of
individuals from different countries, it is physically impossible to assign country of origin.

#3 Sometimes FINANCIALLY the researcher canNOT afford to assign subjects randomly PiggyBankto control & experimental groups.   It costs $ & time to get a list of subjects and then assign them to control & experimental groups using random numbers table or drawing names from a hat.

Thus, researchers sometimes are left with little alternative, but to do a quasi-experiment as the next best thing to an RCT, then discuss its limitations in research reports.

Critical Thinking: You read a research study in which a researcher recruits the 1st 100 patients on a surgical ward January-March quarter as a control group.  Then the researcher recruits the 2nd 100 patients on that same surgical ward April-June for the experimental group.  With the experimental group, the staff uses a new, standardized pain script for better pain communications.  Then the pain communication outcomes of each group are compared statistically.

  • Is this a quasi-experiment or a randomized controlled trial (RCT)?
  • What factors (variables) might be the same among control & experimental groups in this study?
  • What factors (variables) might be different between control & experimental groups that might affect study outcomes?
  • How could you design an ethical & possible RCT that would overcome the problems with this study?
  • Why might you choose to do the study the same way that this researcher did?

For more info: see “Of Mice & Cheese” and/or “Out of Control (Groups).”

Quasi- wha??

Two basic kinds of research design exist:  

  1. Experimental design in which
    • the researcher manipulates some variable,randomized
    • the participants are randomly assigned to groups, &
    • one group is a control group that gets a placebo or some inert treatment so that outcomes in that group can be compared to the group(s) that did get the treatment.
  2. Non-experimental design in which the researcher doesn’t manipulate anything, but just observes & records what is going on.   Some of these are descriptive, correlational, case, or cohort study designs for example.

One particularly interesting “experimental” design is one in which 1 or 2 of the experimental design ideal requirements as listed above are missing.  These are called quasi-experimental designs.

thinking3In a quasi experimental design

  • The researcher manipulates some variable, but….
  • Either the participants are NOT randomly assigned to groups
  • &/OR there is no control group.

A quasi-experimental design is not as strong as a true experiment in showing that the manipulated variable X causes changes in the outcome variable Y.  For example, a true experimental study with manipulation, randomization, and a control group would create much stronger evidence that hospital therapy dogs really reduced patient pain and anxiety.  We would not be as confident in the results of a quasi-experimental design examining the exact same thing.  In the next blog, we’ll examine why.

For more info:  Check out earlier blog:    “What is an RCT anyway?” at https://discoveringyourinnerscientist.com/2015/01/23/whats-a-randomized-controlled-trial/Idea2

Critical thinking:  Go to PubMed & use search terms “experiment AND nurse” (without the quotation marks).  Open an interesting abstract and look for the 3 elements of a classic experimental design. Now look for “quasi experiment AND nurse” (without the quotation marks.)  See what element is missing!

Stand & Deliver: Evidence for Empathy in Action

Patient Pain Satisfaction.  It’s a key outcome of RN empathy in action.CARE

Imagine that you are hospitalized and hurting.   During hourly rounds the RN reassures you with these words:We are going to do everything that we can to help keep your pain under control. Your pain management is our number 1 priority. Given your [condition, history, diagnosis, status], we may not be able to keep your pain level at zero. However, we will work very hard with you to keep you as comfortable as possible.” (Alaloul et al, 2015, p. 323).

Study? In 2015 a set of researchers tested effectiveness of the above pain script using 2 similar medical-surgical units in an academic medical center—1 unit was an experimental unit & 1 was a control unit.  RNs rounded hourly on both units.  handsOn the experimental unit RNs stated the script to patients exactly as written and on room whiteboards posted the script, last pain med & pain scores.  Posters of the script were also posted on the unit.   In contrast, on the control unit RN communication and use of whiteboard were dependent on individual preferences.  Researchers measured effectiveness of the script by collecting HCAHPS scores 2 times before RNs began using the script (a baseline pretest) and then 5 times during and after RNs began using it (a posttest) on both units.

Results? On the experimental units significantly more patients reported that the team was doing everything they could to control pain and that the pain was well-controlled (p≤.05). And while experimental unit scores were trending up, control unit scores trended down. Other findings were that the RNs were satisfied with the script, and that RNs having a BSN or MSN had no effect.

Conclusions/Implications?When nurses used clear and consistent communication with patients in pain, a positive effect was seen in patient satisfaction with pain management over time. This intervention was simple and effective. It could be replicated in a variety of health care organizations.” (p.321) [underline added]

Commentary: While an experiment would have created greater confidence that the script caused the improvements in patient satisfaction, an experiment would have been difficult or impossible.  Researchers could not randomly assign patients to experimental & control units.  Still, quasi-experimental research is relatively strong evidence, but it leaves the door open that something besides the script caused the improvements in HCAHPS scores.

questionCritical thinking? What would prevent you from adopting or adapting this script in your own personal practice tomorrow?  What are the barriers and facilitators to getting other RNs on your unit to adopt this script, including using whiteboards?  Are there any risks to using the script?  What are the risks to NOT using the script?

Want more info? See original reference – Alaloul, F., Williams, K., Myers, J., Jones, K.D., & Logsdon, M.C. (2015).Impact of a script-based communication intervention on patient satisfaction with pain management. Pain Management Nursing, 16(3), 321-327. http://dx.doi.org/10.1016/j.pmn.2014.08.008

“Oh Baby!” Evidence-based Naming Prevents Events

EBP Preventive Action:  Evidence-based, distinct infant naming can avoid sentinel events related to misidentification of newborns (TJC, 2015).

Problem:  Misidentification errors of NICU babies are common newborn3(Gray et al., 2006).   About 12% of the 4 million born in U.S. hospitals were admitted to NICU’s.  At birth every infant requires quick application of an armband, and when parents have not yet decided on a name the assigned name is often quite nondistinct (e.g., BabySmith).

A pretest/posttest of a new, more infant-specific naming system was “conducted in order to examine the effect of a distinct naming convention that incorporates the mother’s first name into the newborn’s first name (e.g., Wendysgirl) on the incidence of wrong-patient errors. We used the Retract-and-Reorder (RAR) tool, an established, automated tool for detecting the outcome of wrong-patient electronic orders. The RAR tool identifies orders placed on a patient that are retracted within 10 minutes and then placed by the same clinician on a different patient within the next 10 minutes” (Adelman et al., 2013). newborn2Their results? RAR events were reduced by 36.3%.   Their recommendations? Switch to a distinct naming system.

Using something like Judysgirl Smith is infant specific. “In the case of multiple births, the hospital adds a number in front of the mother’s first name (ex: 1Judysgirl and 2Judysgirl)” (TJC).

TJC recommends:

  • “Stop using Babyboy or Babygirl as part of the temporary name.
  • Change to a more distinct naming convention.
  • Train staff on the distinct naming convention.
  • Follow the recommendation in National Patient Safety Goal 01.01.01 and implement use of two patient identifiers at all times.
  • As soon as parents decide on their baby’s name, enter that name into the medical record instead of the temporary name.”

Commentary: While this is just one study, RNs should evaluate whether it is riskier to continue any current practice of non-distinct naming or to switch practices to distinct naming. No risks were identified to the distinct naming system & it likely requires only the resource investment of educating staff.  Adelman et al.’s (2013) study is current, moderately strong, quasi-experimental evidence that showed a significant decrease in errors that could have sentinel event outcomes. Any who make the switch should monitor outcomes. All who don’t make the switch should, too!

Critical Thinking: Examine the risks, resources, & readiness of staff in your facility to make the switch to a distinct NICU infant naming system?  question Should the naming system be extended to all infants?

Want more information?  See

Of Mice and Cheese: Research with Non-equivalent Groups

Last week’s blog focused on the strongest types of evidence that you might find when trying to solve a clinical problem. These are: #1 Systematic reviews, Meta-analyses, or Evidence-based clinical practice guidelines based on systematic review of RCTs; & #2 Randomized controlled trials. (For levels of evidence from strongest to weakest, see blog “I like my coffee (and my evidence) strong!”)

So after the two strongest levels of evidence what is the next strongest? #3 level is controlled trials without randomization. (Sometimes called quasi-experimental studies.)

Here’s an example of a controlled trial without randomization: I take two groups of mice and test two types of cheese to find out which one mice like best. I do NOT randomly assign the mice to groups. The experimental group #1 loved Swiss cheese, & the control group #2 refused to eat the cheddar. I assume confidently that mice LOVE Swiss cheese & do NOT like cheddar. What’s the problem with my conclusion? If you want to know, then read on!swiss cheese

In my mouse Controlled Trial Without Randomization, the groups were formed by convenience and Not randomly assigned. Thus, any difference in outcomes between groups might be related to some pre-existing difference between groups. My outcome of mice loving Swiss & hating Cheddar might have nothing to do with the experimental treatment.   In fact, I did not know that all my mice in the Swiss cheese group #1 hadn’t eaten in 2 days, and my mice in the cheddar group #2 had just had a full lunch. Ooops.

On the other hand if I had randomly assigned all the mice to two groups, then I could be relatively confident that all little differences between group members were evenly distributed to both groups, so that the groups were equivalent. My two mouse-groups would have probably ended up with a pretty even distribution of both hungry and not-so-hungry mice.   Then if my Swiss cheese group devoured the Swiss and my cheddar group rejected the cheddar, I could be more certain that mice love Swiss and dislike cheddar.

Happy evidence hunting!