Did your chair write "this reads like an annotated bibliography" in the margin? This guide is for DNP and post-master students who have articles but no synthesis yet. You will learn to search, appraise, tabulate and organize evidence into themes, with tables and worked examples a reviewer can check.
Quick answer. Synthesis means grouping what the studies found, weighing how trustworthy each finding is, and stating what the evidence as a whole supports. A summary reports studies one at a time; a synthesis makes claims across them.
A common committee comment is: "This reads like an annotated bibliography." That comment means the review lists studies rather than combining them. A synthesis has a spine, and the spine is your claim about the evidence.
| Summary style (weak) | Synthesis style (strong) |
|---|---|
| Smith (Year) studied a delirium prevention program on two units. Jones (Year) studied a similar program in one hospital. Lee (Year) reviewed several programs. | Multicomponent programs to prevent delirium in hospitalized older adults were associated with fewer delirium episodes in several studies (Smith, Year; Jones, Year), and a review reached a consistent conclusion (Lee, Year). Effects varied with how fully the components were delivered, which suggests fidelity matters (illustrative structure; cite your real sources). |
The second version leads with the finding, groups sources under it and comments on the pattern. That is the habit to build.
A vague question produces a vague search. Start by rereading your PICOT and splitting it into concepts. Population, intervention and outcome usually become search blocks; comparison and time are more often used in screening than in the search itself.
| Concept | Controlled vocabulary idea | Free-text idea |
|---|---|---|
| Population: hospitalized older adults | Subject headings for aged and inpatients | older adult*, elderly, geriatric, hospitalized |
| Intervention: multicomponent prevention | Subject headings for the intervention type | bundle, multicomponent, non-pharmacologic, protocol |
| Outcome: delirium | The delirium heading in each database | delirium, acute confusion |
The asterisk is a truncation symbol in many databases, so older adult* finds adult and adults. Check each platform's help page because symbols differ.
Use more than one database. Nursing topics commonly draw on CINAHL and MEDLINE (through PubMed or another platform), with Embase, PsycINFO or the Cochrane Library added depending on the subject. The National Library of Medicine maintains the MeSH vocabulary used in MEDLINE.
Keep a search log with database, date, string and number of results. You will need it in an appendix, and reviewers may ask to see it. Add a citation-tracking step, such as checking the reference lists of your best papers, and record what it adds.
The layout below is illustrative and will need adjusting to each database's syntax. It shows how blocks combine.
| Step | Illustrative content |
|---|---|
| Block 1 (population) | "older adult*" OR elderly OR geriatric OR aged |
| Block 2 (intervention) | multicomponent OR bundle OR "non-pharmacologic*" OR protocol |
| Block 3 (outcome) | delirium OR "acute confusion" |
| Combined | Block 1 AND Block 2 AND Block 3, then add subject headings for each block and apply your date and language limits |
Run a quick test: does the search retrieve three or four papers you already know are relevant? If not, the string is missing a synonym or a heading.
Criteria protect you from bias, because they are set before you read the results. Write them as short, testable statements and apply them the same way to every record.
| Criterion | Include | Exclude | Reason to record |
|---|---|---|---|
| Population | Adults in acute care settings | Pediatric and emergency-only samples | Match your project site |
| Intervention | Interventions that resemble your planned change | Single-component drug studies | Relevance to your intervention |
| Design | Systematic reviews, trials, quasi-experimental and improvement studies | Opinion pieces, editorials | Evidence strength |
| Date | A recent range chosen for a reason | Older than the range unless seminal | Currency of practice |
| Language and access | English full text | Abstract-only records | Feasibility, stated as a limitation |
Justify every limit. A date cutoff of five years is common but not a rule. Say why you chose it, and add a line about seminal older studies if any are essential.
Screen in stages: titles and abstracts first, then full texts. Note the reason for each full-text exclusion. Many students report their numbers in a flow diagram modeled on the PRISMA 2020 flow diagram, which you can find through the EQUATOR Network. PRISMA is written for systematic reviews, so adopt it as a model and state that your synthesis is a focused review, not a full systematic review. Our systematic review guide covers the stricter version.
Appraisal answers two questions: how well was the study done, and how much weight should its findings carry? Match the tool to the design, and say which you used.
| Evidence type | Examples of appraisal tools | Notes |
|---|---|---|
| Systematic reviews | AMSTAR 2, JBI checklist | Assess how the review itself was conducted |
| Randomized trials | Cochrane risk of bias tool (RoB 2), JBI checklist | Look at randomization, blinding, missing data |
| Non-randomized studies | ROBINS-I, JBI checklists, Newcastle-Ottawa Scale | Confounding is the main concern |
| Qualitative studies | JBI checklist, CASP | Credibility, congruity of method and interpretation |
| Clinical practice guidelines | AGREE II | Development rigor and applicability |
| Quality improvement studies | SQUIRE 2.0 as a reporting reference | Judge completeness and design honesty |
The JBI critical appraisal tools cover many designs and are widely used in nursing. Some programs also require a level-of-evidence system such as the Johns Hopkins Evidence Level and Quality Guide or the Melnyk hierarchy. Use the one your program names, and apply it consistently. Where you appraise the body of evidence as a whole, the GRADE approach is a common option.
You do not need to reproduce every checklist item. State the tool, the overall judgment for each study, and the reasons for any concern. A short phrase such as "moderate quality due to no comparison group and a small single-site sample" tells the reader more than a score alone.
The evidence table is your working document and usually an appendix. It lets a reader check your reading in a minute. Keep columns consistent across studies, and keep cell text brief.
| Column | What to enter |
|---|---|
| Citation | APA 7 reference or short author-year form |
| Purpose and design | Aim, design, setting |
| Sample | Size and key characteristics |
| Intervention and comparison | What was done, versus what |
| Measures | Main outcomes and how measured |
| Key findings | Results in your own words, with the direction and size of effect if reported |
| Limitations | Threats to validity noted by authors and by you |
| Level and quality | Result from the appraisal tool |
For formatting help see the synthesis matrix resource and the APA tables and figures guide.
Send your PICOT, search terms and articles with your brief, and we can build or edit the evidence table and the synthesis narrative. The price is shown before you pay, and every delivered paper includes 14 days of free revisions.
A reference manager stores your PDFs and citations, and a simple spreadsheet holds the evidence table. Name files consistently, for example author, year and topic, and keep a "reasons for exclusion" column so your flow numbers are easy to rebuild.
If two people can screen a sample of records independently, note your agreement and how you resolved disagreements. Even a small check shows the committee that screening was not a solo judgment call.
Synthesis begins when you stop asking "what did this study find?" and start asking "what do these studies say together?" Read your evidence table column by column instead of row by row.
| Approach | Best when | Watch for |
|---|---|---|
| Narrative synthesis | Studies differ in design and measures | Needs explicit rules for how findings were grouped |
| Thematic synthesis | Qualitative or mixed-method evidence | Keep a clear line from data to theme |
| Meta-analysis | Similar studies report comparable numeric outcomes | Usually beyond a DNP synthesis unless already published |
Conflict between studies is not a problem to hide. It is information. Look for a reason the studies might differ before you conclude that the evidence is contradictory.
| Possible reason | What to check | How to write it |
|---|---|---|
| Different populations | Age, acuity, setting | "Findings differed between surgical and medical samples..." |
| Different intervention intensity | Components, dose, delivery | "Programs delivering fewer components reported smaller effects..." |
| Different outcome definitions | Tools, timing, thresholds | "Outcomes were measured with different instruments..." |
| Different study quality | Design, bias risk | "The two higher quality studies agreed that..." |
Each paragraph should carry one claim about the evidence. A reliable pattern is claim, support, comment, and link.
Avoid opening paragraphs with an author's name. When the author is the subject, you are usually summarizing. Follow APA 7 for citations; our APA in-text guide covers multiple sources in one bracket.
| Weak topic sentence | Strong topic sentence |
|---|---|
| Several studies looked at staff education. | Staff education alone produced inconsistent gains, while education combined with reminders was more consistently associated with practice change (cite sources). |
| Research on barriers has been done. | Time pressure and unclear ownership were the barriers most often reported across implementation studies (cite sources). |
The gap statement is the bridge to your project. It is not "more research is needed". It is a specific statement about what the evidence does not answer for your setting and how your project responds.
| Type of gap | Example wording pattern |
|---|---|
| Population gap | Most studies were conducted in specialist units; few describe general medical wards like ours. |
| Implementation gap | Effectiveness is reported, but few studies describe how staff adopted the components. |
| Measurement gap | Outcomes were defined differently across studies, limiting comparison. |
| Sustainability gap | Follow-up beyond the intervention period is rarely reported. |
Link the gap forward to your aims and design. Your methodology should visibly respond to it.
Invented for structure only, this shows how a synthesis section can be organized. It contains no real study results.
| Theme | Claim (illustrative) | Evidence quality note | Implication for the project |
|---|---|---|---|
| Intervention components | Programs combining several non-drug components appear to be favored over single components. | Mix of reviews and quasi-experimental studies | Adopt a bundle, not a single tool |
| Implementation fidelity | Benefits appear larger where components were delivered consistently. | Mostly observational | Plan fidelity monitoring |
| Staff engagement | Programs with unit champions report better uptake. | Small studies | Recruit a champion from each shift |
| Measurement | Outcome tools varied widely. | Not applicable | Choose one validated tool and define it clearly |
| Feedback | What it means | Fix |
|---|---|---|
| "Where is the appraisal?" | Quality is assumed rather than judged. | Add a tool name and a quality column to the evidence table |
| "This is dated." | Too many old sources, or no recent ones. | Re-run the search with a defended date range |
| "Too descriptive." | Studies are reported in sequence. | Reorganize under theme headings with one claim each |
| "How does this change your project?" | No link between evidence and design. | End each theme with an implication statement |
Treat these as relative proportions rather than fixed durations, because your search size and other coursework will shift them.
Illustrative example, not a real client. This short story is invented to show the pattern, and it contains no real people or numbers.
The problem. A student had a folder full of articles and wrote them up one by one, each in its own paragraph.
The tension. Her chair wrote "annotated bibliography" in the margin, and she feared starting over.
The turn. She built an evidence table with a quality column, sorted the findings into four themes and rewrote each section around one claim per theme.
The proof. The chair's next comment was "now I can see your argument", and the committee accepted her gap statement.
The payoff. The four themes later became the structure of her intervention plan, so the review did real work in the project.
There is no universal number. Your program may set a range, but the better test is whether you have enough good evidence to justify the intervention and describe its limits.
Sometimes. Guidelines, reports and dissertations can add value, especially for implementation detail, but appraise them carefully and label them clearly. See the grey literature citation guide.
Only if your program requires a systematic review. For a focused DNP synthesis, many programs accept a PRISMA-style flow diagram and a clear search log.
Say so plainly. A synthesis that honestly describes weak evidence and explains how your design compensates is stronger than one that overstates certainty.
You can rely on a high-quality recent review for the broad picture, but then update it with newer primary studies and explain the approach. Ask your chair whether that is acceptable.
Request them through your library's interlibrary loan service, or contact the author. If a key article stays unavailable, list it as a limitation instead of citing it from the abstract alone.
A good synthesis ends with a decision: adopt, adapt or reconsider the intervention, and here is why. Keep your search reproducible, your appraisal explicit and your paragraphs organized around claims. That structure is what convinces a committee that your project rests on more than a hunch.
Want the evidence table or synthesis built or edited? Get my instant quote. The price is shown before you pay, every delivered paper includes 14 days of free revisions, and refund terms are on the money-back guarantee page. Please use any model paper in line with your institution's academic-integrity policy.