Nursing Dissertation Guide

How to Solve Common Evidence-Based Practice Dissertation Problems

Diagnose and repair evidence-based practice dissertation problems across question design, searching, appraisal, synthesis, context, recommendations and evaluation.

Written by Moses Updated August 25, 2026 8 min read
How to Solve Common Evidence-Based Practice Dissertation Problems

Evidence-based practice dissertation problems are often alignment problems rather than writing problems. A dissertation can contain many sources yet remain weak if the practice concern, research question, search strategy, appraisal, synthesis and recommendation do not form one defensible chain.

This guide shows how to locate the broken link and repair it without adding unnecessary material or overstating the evidence.

Use an evidence-to-decision chain

  1. Define a meaningful nursing practice problem.
  2. Convert it into an answerable question.
  3. Search transparently for relevant evidence.
  4. Appraise each study according to its design and purpose.
  5. Synthesise findings rather than listing papers.
  6. Interpret evidence in relation to patients, context and feasibility.
  7. Keep recommendations proportionate to certainty.
  8. Separate implementation from evaluation.

Problem 1: the practice problem is asserted rather than demonstrated

Statements such as “this is a major problem” need evidence. Build the introduction in three layers:

  1. Show the clinical or service burden with credible evidence.
  2. Explain why the problem matters to nursing practice.
  3. Identify the unresolved decision or evidence gap the dissertation will address.

National statistics do not automatically prove a local service problem. If local audit data are unavailable, state that limitation and avoid fabricating a baseline.

Problem 2: the review question does not match the real decision

A technically correct PICO question can still be unhelpful if it does not represent the actual nursing decision.

For example, “Does oral care prevent pneumonia?” is too broad. A more focused effectiveness question might ask whether a defined nurse-led oral-care bundle reduces non-ventilator hospital-acquired pneumonia among older adult inpatients compared with usual care.

If the real uncertainty concerns acceptability for people with dementia, however, a qualitative experience question may be more appropriate.

Write one sentence explaining what decision the review is intended to inform. Then check whether every eligibility criterion, synthesis heading and recommendation still serves that decision.

Problem 3: eligibility criteria change after seeing the evidence

Changing criteria to keep an attractive paper introduces selection bias. Define population, intervention or phenomenon, comparator where relevant, outcomes, setting, design, publication type and justified date or language limits before screening.

If the criteria genuinely need to change, document what changed, when and why. Transparent amendment is stronger than pretending the revised rule was planned from the beginning.

Problem 4: the search is broad but not reproducible

Naming databases is not enough. Readers should be able to see how question concepts became search terms, subject headings and Boolean combinations.

Record:

  • database and platform;
  • complete search strategy;
  • search date;
  • limits and their rationale;
  • result counts; and
  • deduplication method.

PRISMA 2020 supports transparent systematic-review reporting, while the site’s systematic-review search-strategy guide covers the practical search process (Page et al., 2021).

Problem 5: the evidence hierarchy is used mechanically

Study design should be judged against the claim. Randomised trials are useful for many intervention-effect questions but do not automatically answer questions about acceptability, implementation or lived experience.

Qualitative research may provide the most relevant evidence for how patients or nurses experience an intervention. Observational designs can be appropriate when randomisation is infeasible or unethical.

Do not call qualitative evidence “low level” and then rely on it to explain why an intervention is acceptable. Match the design to the component of the decision it can inform.

Problem 6: critical appraisal becomes a checklist score

Appraisal tools structure judgement; they do not replace it. CASP provides design-specific appraisal checklists, but recording yes/no answers is only the first step (Critical Appraisal Skills Programme [CASP], n.d.).

For each important limitation, explain the consequence:

Because attrition was high and differed between groups, confidence in the long-term comparison is reduced because the analysed participants may not represent those originally allocated.

Then look for recurring limitations across the evidence base, such as reliance on self-report, short follow-up or specialist settings.

Problem 7: the results chapter summarises papers one by one

A sequence of author-by-author summaries forces the reader to perform the synthesis. Organise findings around outcomes, intervention components, experiences or explanatory themes instead.

Use an evidence table to compare:

  • design and sample;
  • setting;
  • intervention or phenomenon;
  • outcomes or themes;
  • follow-up; and
  • important limitations.

The nursing evidence-table guide provides a structured extraction method.

Problem 8: heterogeneity is mistaken for contradiction

Different findings may reflect genuine differences in population, setting, intervention intensity, measurement or follow-up rather than a simple disagreement.

Compare Why it matters
Population risk Baseline risk can change the likely benefit.
Setting Specialist and general services may not be comparable.
Intervention components Education alone is not equivalent to education plus audit and feedback.
Outcome definition Knowledge, adherence and clinical outcomes answer different questions.
Follow-up Immediate change may not persist.

Map these differences before deciding that the literature is inconsistent.

Problem 9: statistical significance is confused with certainty

A significant p-value does not automatically mean the result is large, precise, clinically important or trustworthy. A non-significant result does not prove that no effect exists.

Consider effect size, confidence intervals, risk of bias, consistency, directness and precision. GRADE explicitly separates certainty of evidence from the strength of a recommendation (GRADE Working Group, n.d.).

Use calibrated language such as suggests, may improve or evidence remains uncertain when the evidence does not justify stronger wording.

Problem 10: context appears only in the conclusion

Evidence does not implement itself. Staffing, workflow, resources, leadership, patient preferences, accessibility and health inequalities can change whether an intervention is feasible or meaningful.

The JBI Model of Evidence-Based Healthcare highlights feasibility, appropriateness, meaningfulness and effectiveness as important dimensions of evidence-informed decision-making (JBI, 2016).

For a nurse-led oral-care bundle, for example, the dissertation may need to consider product availability, staff competence, patient consent, swallowing risk, cognitive impairment and documentation burden—not only whether studies reported an effect.

Problem 11: recommendations exceed the evidence

A recommendation should be the final result of the reasoning chain.

Weak:

All hospitals should implement the bundle immediately.

More proportionate:

Services with a documented oral-care gap could consider a governed pilot of the multicomponent bundle, with staff preparation, process measurement and monitoring for unintended burden.

The second recommendation reflects uncertainty and acknowledges the need for local evaluation.

Problem 12: implementation and evaluation are merged

Implementation explains how a proposed change would be introduced. Evaluation explains how delivery and effects would be measured.

Implementation may cover stakeholders, responsibilities, resources, training, governance and phased rollout. Evaluation should identify baseline, process measures, patient or clinical outcomes, balancing measures, data sources and review timing.

If the dissertation is literature-based, present these as proposals. Do not invent local results or claim that implementation has occurred.

Worked repair example

Imagine a dissertation titled “Oral care and infection.” Its search includes paediatric dental clinics, ventilated ICU patients, care homes and general adult wards, yet the conclusion recommends chlorhexidine oral care for every adult inpatient.

A stronger repair would:

  1. define the real decision, such as structured nurse-led oral care for non-ventilated older adults on acute medical wards;
  2. align the eligibility criteria with that population and setting;
  3. treat ICU evidence as indirect where appropriate;
  4. include qualitative evidence where acceptability matters;
  5. group findings by clinical outcomes, oral-health outcomes, delivery reliability and acceptability; and
  6. recommend a local pilot only if the evidence and baseline need justify it.

The repaired project becomes narrower but more defensible.

Final EBP chain audit

  • The introduction demonstrates a specific nursing-practice problem.
  • The review question expresses a real evidence or practice decision.
  • Eligibility criteria are justified and consistently applied.
  • The search is reproducible.
  • Study designs are judged by fitness for the question.
  • Appraisal explains consequences rather than checkbox results.
  • Synthesis compares evidence by outcomes, themes or mechanisms.
  • Heterogeneity is examined before findings are called contradictory.
  • Effect magnitude and certainty are considered alongside significance.
  • Patient preferences, equity and feasibility shape interpretation.
  • Recommendations remain proportionate.
  • Implementation and evaluation are kept distinct.
  • The conclusion answers the original question.

Frequently asked questions

Is an EBP dissertation the same as a systematic review?

No. Some EBP dissertations use systematic-review methods, while others use integrative, structured or rapid evidence approaches. Use the method required by the programme and label it accurately.

Can qualitative and quantitative studies be included?

Yes, when both are needed to answer the question and the method explains how they will be appraised and synthesised.

What if the evidence is weak?

Explain the uncertainty, identify what can still be inferred and make cautious recommendations rather than forcing a strong conclusion.

How many studies should be included?

There is no universal number. Transparent selection, relevance and analytical depth matter more than reaching an arbitrary target.

Should every EBP question use PICO?

No. PICO suits many intervention-effect questions, but other frameworks may better fit experience, exposure, policy or service questions.

Project-specific support

If an existing EBP draft has lost alignment between the question, evidence table and recommendations, use the nursing EBP project support route or the contact page.

Related Nursing Guides

Conclusion

Solving evidence-based practice dissertation problems requires tracing the whole evidence-to-decision chain. Define the decision, align the question and search, appraise the consequences of methodological limitations, explain heterogeneity, judge certainty and keep recommendations proportionate. When every link supports the next, the dissertation becomes clearer and more defensible.

References

Related articles

Browse all