EBP, research and QI, plus the projects that apply them

Which project would require review by an institutional review board (IRB)?

1. A PDSA test of a new handoff tool on a single unit
   A PDSA test of a handoff tool on a single unit is local quality improvement, aimed at improving one process rather than producing generalizable knowledge.
2. Randomizing nurses to two teaching methods
   Randomizing nurses to two teaching methods is designed to produce generalizable knowledge, which makes it human subjects research that needs IRB review.
3. An audit of compliance with the hand hygiene policy
   An audit of compliance with the hand hygiene policy measures current practice for internal improvement and is not research.
4. A QI project the team plans to present at a conference
   A QI project the team plans to present at a conference stays QI; the intent to share results does not turn a local improvement project into research.

A pilot study with 12 nurses per group found no significant difference in knowledge scores between e-learning and classroom teaching. What is the main concern in interpreting this result?

1. A Type II error from low power
   A Type II error from low power is the main concern: with 12 nurses per group, the study may miss a real difference and report a false negative.
2. It proves the two methods are equivalent
   Saying it proves the two methods are equivalent mistakes absence of evidence for evidence of no difference.
3. A Type I error from testing many outcomes
   A Type I error from testing many outcomes is a false positive, the opposite of a nonsignificant result.
4. Selection bias created by randomization
   Selection bias created by randomization reverses the logic, since randomization is what reduces selection bias.

An NPD practitioner plans an anonymous, minimal-risk survey of nurses' scheduling preferences for a research study. What review is most likely needed?

1. Expedited review by the IRB
   Expedited review by the IRB covers minimal-risk research that does not fit an exempt category; an anonymous preference survey usually does fit one.
2. An exempt determination by the IRB
   An exempt determination by the IRB is the likely route: anonymous, minimal-risk surveys usually fit an exempt category, but the IRB or its office makes that call.
3. None; the investigator self-declares it exempt
   Having the investigator self-declare it exempt skips the required independent determination; the person running the study has a conflict of interest.
4. Full board review by the IRB
   Full board review by the IRB is reserved for research above minimal risk, which this anonymous survey is not.

A unit implemented a checklist that reduced catheter-associated urinary tract infections. Which strategy most effectively sustains the improvement?

1. A staff attestation of understanding
   A staff attestation of understanding documents that staff read something, not that they keep using the checklist.
2. Ongoing process audits with feedback
   Ongoing process audits with feedback measure checklist use and show staff the results, so drift is caught before infections rise again.
3. Monitoring infection rates only
   Monitoring infection rates only tracks a lagging outcome, so the process can slip for months before infections reveal it.
4. An annual refresher module for staff
   An annual refresher module for staff treats sustainment as a knowledge problem and leaves a year between checks.

When translating research into practice, why is a confidence interval generally more useful to the NPD practitioner than a p-value alone?

1. It shows the size and precision of the effect
   It shows the size and precision of the effect, which helps judge clinical importance, while a p-value alone only addresses chance.
2. It eliminates the chance of a Type I error
   Eliminating the chance of a Type I error is impossible; a confidence interval carries the same error rate as the matching significance test.
3. It confirms the statistical significance of results
   Confirming the statistical significance of results is what the p-value already does, so it is not the added value of the interval.
4. It establishes the internal validity of the study
   Establishing the internal validity of the study depends on design and bias control, which no statistic can provide.

A phenomenological study of nurse resilience recruits participants only through a flyer in the breakroom. Which sampling approach would better fit this design?

1. Stratified random sampling
   Stratified random sampling is a probability method for representing subgroups in quantitative research, not a qualitative strategy.
2. Random sampling
   Random sampling aims at statistical representativeness for generalization, which a phenomenological study does not seek.
3. Purposive sampling
   Purposive sampling deliberately selects people who have lived the experience and can describe it, which is what phenomenology needs.
4. Quota sampling
   Quota sampling fills preset demographic counts, which does not ensure that participants have rich experience of resilience.

An intervention reduces the fall rate from 4% to 2% (p < 0.01). Which metric most helps leadership make resource allocation decisions?

1. The p-value (less than 0.01)
   The p-value shows the effect is unlikely to be chance, not how large it is, so it cannot guide resource decisions.
2. Relative risk reduction (50%)
   Relative risk reduction (50%) sounds larger but hides how common falls are, so it can overstate the practical gain.
3. The odds ratio for falls
   The odds ratio for falls is harder to interpret for planning and drifts from the actual risk difference.
4. Absolute risk reduction (2%)
   Absolute risk reduction, 2 fewer falls per 100 patients, shows the real-world size of the benefit and leads to a number needed to treat that leaders can cost out.

A qualitative researcher keeps a detailed audit trail of every methodological decision and change in data collection. This practice primarily strengthens which trustworthiness criterion?

1. Dependability
   Dependability is the consistency and traceability of the research process, and an audit trail lets an outside reviewer follow every decision.
2. Credibility
   Credibility, the truth value of findings, is mainly strengthened by member checking and prolonged engagement, not by documenting decisions.
3. Generalizability
   Generalizability is a quantitative criterion; qualitative research uses trustworthiness criteria instead.
4. Transferability
   Transferability depends on thick description that lets readers judge fit to other settings, which an audit trail does not supply.

In a fall prevention pilot, the intervention unit (n = 100) had 5 falls and the control unit (n = 100) had 10 falls. What is the relative risk of falling on the intervention unit?

1. 5.0
   5.0 confuses the count of falls with a ratio of risks.
2. 0.5
   0.5 is the relative risk: 5/100 (0.05) on the intervention unit divided by 10/100 (0.10) on the control unit.
3. 0.05
   0.05 is the intervention unit's absolute risk, not the ratio between the two groups.
4. 2.0
   2.0 divides control risk by intervention risk, inverting the ratio.

A proposed EHR update reduces documentation time by 2 seconds per note (p = 0.03). The predetermined threshold for practical significance was a 15-second reduction. What is the most appropriate recommendation?

1. Implement it as a pilot project first
   Implementing it as a pilot project first spends resources testing a change already shown to fall short of the practical threshold.
2. Implement due to statistical significance
   Implementing due to statistical significance treats a small p-value as proof of importance, which it is not.
3. Do not implement the change
   Not implementing the change is right, because 2 seconds falls far short of the 15-second threshold for practical significance set in advance.
4. Conduct a study with a larger sample
   Conducting a study with a larger sample would only make a trivial effect more precise; the effect is already known to be too small.

The NPD practitioner plans to evaluate a new mentorship program's impact on nurse turnover by measuring retention one year after implementation. What is the primary flaw in this plan?

1. Retention is not a valid outcome
   Retention is not a valid outcome is wrong, since retention is a central goal of mentorship programs.
2. Qualitative data are also required
   Saying qualitative data are also required adds a nice extra but does not fix the missing comparison.
3. No baseline retention data
   No baseline retention data is the flaw, because impact means change, and post-program retention needs a before or comparison figure.
4. A one-year window is too short
   A one-year window being too short is not the problem; first-year retention is a standard window for this outcome.

A sepsis improvement project reports 98% adherence to the new bundle checklist, yet 30-day mortality is unchanged. Which conclusion is most supported by these data?

1. Insufficient staff training
   Insufficient staff training would show up as poor adherence, not as near-perfect adherence without an outcome change.
2. Implementation failure
   Implementation failure means the intervention was not delivered as planned, which 98% adherence rules out.
3. Sustainment failure
   Sustainment failure means gains faded over time, but there was no gain to lose here.
4. Program theory failure
   Program theory failure fits: the bundle was delivered as intended, yet the outcome did not change, so the assumption linking intervention to outcome is in question.

A PDSA test of a new handoff tool shows better handoff satisfaction but more overtime because the tool takes too long to complete. Which Act-phase decision is most appropriate?

1. Adapt the tool and retest it
   Adapting the tool and retesting it keeps the satisfaction gain while fixing the time burden, then checks the change in another cycle.
2. Abandon the tool because of cost
   Abandoning the tool because of cost throws away a real improvement that a shorter version might keep.
3. Repeat the same test unchanged
   Repeating the same test unchanged would reproduce the known overtime problem without learning anything new.
4. Adopt the tool on all units
   Adopting the tool on all units would spread the overtime problem to every unit.

An NPD practitioner shares PDSA documentation from a pilot unit with units that will adopt the practice next. Which information is most useful to them?

1. The pilot unit's staffing levels
   The pilot unit's staffing levels are context, but alone they do not guide how to adapt the practice.
2. The pilot's final outcome data
   The pilot's final outcome data build confidence but do not show how the practice was adapted along the way.
3. Statistical significance of results
   Statistical significance of results is rarely the point of PDSA, which uses small, rapid tests rather than formal hypothesis testing.
4. What changed in each cycle and why
   What changed in each cycle and why shows receiving units how the practice was adapted, so they can make their own adjustments.

The NPD practitioner secures an executive sponsor for a system-wide practice change. What is the primary role of this sponsor?

1. Deliver the education sessions to frontline nurses
   Delivering the education sessions to frontline nurses is the work of NPD practitioners and other educators.
2. Manage the project's daily task schedule and budget
   Managing the project's daily task schedule and budget is the project manager's job, not the senior sponsor's.
3. Provide authority and resources and remove barriers
   Providing authority and resources and removing barriers is the executive sponsor's role, especially for obstacles that cross departments.
4. Model the new practice as a unit peer champion
   Modeling the new practice as a unit peer champion is the role of frontline champions, not an executive.