These 60 selected concepts support preparation for the National Dental Specialty Examination (NDSE) in Dental Public Health. Start with population measures, then apply them to community problems, policy choices and program decisions. Each concept includes an original worked example and a specific error to avoid. Read the cited references alongside the concepts, especially when interpreting Canadian coverage policies or research requirements.
Community assessment and population measures
1. Prevalence depends on the observation window
Point prevalence counts all existing cases, new and pre-existing, on one specified date. Period prevalence counts everyone affected at any time during a defined interval. The denominator must match the population and period described, and the illustrative population is assumed stable.
Worked example: In a stable population of 200 residents, 30 have dental anxiety on the survey date, giving 15% point prevalence; 50 distinct residents were affected at some point during the year, giving 25% period prevalence.
Mistake to avoid: Counting each repeated observation of the same person as a separate case in the period prevalence numerator.
Reference: [4]
2. Incidence proportion measures risk over a stated period
Incidence proportion, or risk, is the proportion of an initially disease-free population that develops the stated outcome during a specified period. The denominator must be limited to the population at risk, and the example assumes complete follow-up of persons, not teeth or surfaces.
Worked example: Among 300 children initially free of defined dental trauma, complete one-year follow-up identifies 24 new person-level cases, so the one-year incidence proportion of that trauma outcome is 24 divided by 300, or 8%.
Mistake to avoid: Leaving children who already had the condition in the starting denominator of the incidence proportion.
Reference: [4]
3. Person-time accommodates unequal follow-up
An incidence rate divides new cases by the summed time each disease-free person was observed, stopping at onset, death, loss to follow-up or study end. It accommodates participants entering and leaving, but pooling person-time assumes risk does not vary over time, so the rate is not automatically an individual probability.
Worked example: In a cohort where participants contribute unequal observation periods, six new cases of a defined condition arise during 150 summed person-years, yielding 0.04 cases per person-year, reported as 40 per 1,000 person-years.
Mistake to avoid: Reporting the person-time rate as a 4% one-year probability despite participants contributing different observation periods.
Reference: [4]
4. Prevalence reflects both occurrence and duration
Prevalence depends on both incidence and illness duration. High prevalence may reflect frequent new cases, prolonged duration without cure, or both, so prevalence alone cannot establish incidence. Changes in recovery or survival can shift prevalence without any corresponding increase in new disease.
Worked example: Two hypothetical communities each record ten new cases yearly, but one retains more unresolved cases, so its higher prevalence alone does not establish higher incidence; this interpretation exercise is not a quantitative prevalence model.
Mistake to avoid: Interpreting every observed increase in prevalence as proof that new disease occurrence has accelerated.
Reference: [4]
5. Combine complementary assessment evidence
A situational assessment relies on many types and sources of data and collection methods, so health indicators, community accounts, evaluation findings and guidance can each answer different assessment questions when combined to inform planning.
Worked example: A survey records which residents missed dental visits, interviewed residents report transport difficulties as their explanation, and service records list available appointment times, together describing access concerns and resources.
Mistake to avoid: Treating one resident's interview account as a numerical estimate of community-wide prevalence.
Reference: [6]
6. Assess community assets as well as unmet needs
A situational assessment should meaningfully engage the intended audience and highlight community assets, not only deficits and problems, while any identified asset still requires confirmation of its availability, suitability and willingness to participate.
Worked example: For a proposed oral health education program, residents suggest an accessible community centre and an existing volunteer network, and planners confirm the centre's availability and the volunteers' willingness before mapping these resources.
Mistake to avoid: Producing a deficit-only report that overlooks resources already available locally in the community.
Reference: [6]
Understanding community problems and determinants
7. Risk ratios compare probabilities
A risk ratio divides the risk, such as the incidence proportion or attack rate, in the group of primary interest by the comparison group's risk, using comparable outcomes and observation periods.
Worked example: In a workplace cohort, 20 of 100 unvaccinated staff and 10 of 100 vaccinated staff develop influenza over the same season, giving a risk ratio of 0.20 divided by 0.10, or 2.
Mistake to avoid: Dividing 20 cases by 10 cases directly when the two group sizes differ.
Reference: [5]
8. Rate ratios retain person-time denominators
A rate ratio divides the incidence or person-time rate of the group of primary interest by the comparison group's rate, after converting both to the same time units and scaling factor.
Worked example: A clinic records 12 periodontal referrals over 600 patient-months and 8 over 800 patient-months; the rates are 20 and 10 per 1,000 patient-months, so the rate ratio is 2.
Mistake to avoid: Substituting enrolled participant counts for person-time when follow-up lengths differ between the groups.
Reference: [5]
9. Odds ratios use the four cells of a comparison
With exposure forming the rows and disease the columns, the odds ratio is the cross-product ad/bc. Define exposure, outcome and reference categories first, then multiply the concordant cells and divide by the product of the discordant cells.
Worked example: In a hypothetical complete 2×2 comparison with no zero cells, exposed cases number 30, exposed non-cases 70, unexposed cases 15 and unexposed non-cases 85, giving OR = (30×85)/(70×15) = 2.43.
Mistake to avoid: Swapping one cell, such as using 70 instead of 30, without changing the stated comparison.
Reference: [5]
10. Case-control sampling does not supply population risks
In a typical case-control study, the investigator decides how many controls to enrol and the source population size is often unknown, so risks, rates, risk ratios and rate ratios cannot be calculated; the odds ratio remains the available measure.
Worked example: Investigators enrol 60 cases and 120 controls; the fraction 60/180 = 33.3% describes only the selected sample's composition, not community disease prevalence, because control numbers were chosen by the investigators.
Mistake to avoid: Dividing selected cases by all participants and reporting it as community disease prevalence.
Reference: [5]
11. Odds ratios are not generally risk ratios
When the health outcome is uncommon, the odds ratio reasonably approximates the risk ratio, and the more common the disease, the further the odds ratio drifts from the risk ratio; the cited passage specifies no numerical rarity threshold.
Worked example: Using hypothetical complete population data, risks of 40% and 20% give RR = 2, while the corresponding odds give OR = (0.4/0.6)/(0.2/0.8) = 2.67, showing divergence for a common outcome.
Mistake to avoid: Interpreting an odds ratio of 2.67 as meaning 2.67 times the risk.
Reference: [5]
12. Diagnose determinants beyond individual behaviour
Health promotion addresses root causes of poor health, including the structural, social, economic and ecological factors shaping health, so assessment should consider these determinants alongside community assets and the local planning context, not personal behaviour alone.
Worked example: Residents report missed appointments because transport is unavailable; assessment should examine the access environment, alongside community assets and local planning context, rather than assume residents lack knowledge about attending.
Mistake to avoid: Selecting an education-only intervention before examining the documented transport barriers.
Reference: [6]
13. Interpret Canadian access disparities in their period
The pre-CDCP Canadian report describes income-related differences in dental access as a baseline context for 2018–2024, not an evaluation of CDCP effects, because the plan was not yet available to help pay for oral health care.
Worked example: In the report, cost-related avoidance of dental visits among people with family incomes under $90,000 rose from 29% to 34% between 2018 and 2024, a five-percentage-point absolute increase.
Mistake to avoid: Attributing that reported five-point rise to the later availability of CDCP coverage.
Reference: [13]
14. Service use and lived oral health burden are different measures
A community needs assessment can distinguish delayed service use, such as gaps since the last dental visit, from lived oral health burden, such as self-reported persistent mouth pain, because these indicators measure different things.
Worked example: In a hypothetical survey of 200 adults, 80 report no dental visit for over a year (40%) while 30 report persistent mouth pain (15%), which are separate indicators.
Mistake to avoid: Treating an adult's lack of a recent dental visit as a clinical disease diagnosis.
Reference: [13]
15. Separate observed association from causal attribution
A comparison can show an association between service use and an outcome without establishing why it occurred, because the evaluation design determines which claims about causation the evidence can support.
Worked example: A one-time survey finds less reported mouth pain among people who used dental services, but this observational comparison does not establish that service use caused the difference.
Mistake to avoid: Presenting a noncausal observational comparison as proof of a program's effectiveness.
Reference: [7]
Canadian oral health policy and financial analysis
16. CDCP enrolment does not guarantee every service
Under the CDCP Dental Benefits Guide effective April 1, 2026, coverage of oral health care services is determined on an individual basis considering criteria such as the client's oral health status and medical conditions, so providers must confirm coverage.
Worked example: Two enrolled CDCP clients request the same restorative service; each client's oral health status, medical conditions and applicable program criteria still require separate individual assessment before coverage.
Mistake to avoid: Promising a client payment for a service solely because they are CDCP-enrolled.
Reference: [12]
17. Distinguish exclusions from frequency limits
Under the April 1, 2026 CDCP guide, an exclusion is always outside coverage, while certain eligible services above frequency limits may be considered through preauthorization, which is service-specific and does not guarantee approval.
Worked example: A client with a physical disability who needs support brushing may have extra eligible hygiene services considered via preauthorization, but an implant stays excluded regardless of any request submitted.
Mistake to avoid: Treating every coverage restriction, including implant exclusion, as an overridable frequency limit.
Reference: [12]
18. Preauthorization and urgent post-determination differ
For CDCP services requiring preauthorization, approval must come before treatment begins; post-determination is intended rarely, only for emergent clinical situations, requires the same documentation plus a rationale, and applies the same policies without guaranteeing payment.
Worked example: A hypothetical service stipulated to require preauthorization follows advance approval before treatment; if retrospective review is sought for an emergent clinical situation, the request needs that service's preauthorization documentation plus a rationale.
Mistake to avoid: Routinely using retrospective post-determination review to bypass the advance preauthorization approval requirement.
Reference: [12]
19. Rolling coverage periods do not reset annually
CDCP frequency limitations run on rolling periods, not calendar years. Eligibility for the next service begins the day after the stated period ends, so January 1 does not restore a used allowance.
Worked example: For a hypothetical service limited to one in any 12 months delivered on July 10, 2026, the guide's illustrated rule places the next eligibility on July 11, 2027, not January 1, 2027.
Mistake to avoid: Assuming the January 1 calendar year resets the client's service allowance.
Reference: [12]
20. Coverage decisions require evidence of the criteria
CDCP preauthorization submissions must demonstrate the requested service's eligibility criteria, with documentation requirements varying by procedure and each client's needs and circumstances; insufficient submissions may prompt requests for more information or denial, though some incomplete submissions are considered.
Worked example: A hypothetical preauthorization request omits information needed to demonstrate one relevant eligibility criterion, so the CDCP may request additional information or deny the submission based on the information provided.
Mistake to avoid: Assuming a submitted form alone proves entitlement, or that every incomplete submission must be rejected.
Reference: [12]
21. Approval validity and client coverage are separate checks
A CDCP preauthorization decision's validity period and the client's coverage on the service date are separate conditions: most approvals last up to 12 months, some preventive and periodontal ones up to 24 months, yet payment still requires current eligibility on the service date.
Worked example: A hygienist confirms a periodontal preauthorization approved ten months ago still falls within its validity window, then verifies the client's CDCP enrolment remains active for the planned service date before treatment proceeds.
Mistake to avoid: Assuming an approval still inside its validity period guarantees the client's enrolment remains active.
Reference: [12]
22. Fluoridation targets and maximum concentrations differ
The guideline sets a 1.5 mg/L maximum acceptable concentration for fluoride in drinking water, while the optimal concentration for dental health in communities that choose to fluoridate is 0.7 mg/L; these figures serve different purposes and should not be conflated.
Worked example: A public health dentist reviewing a town's water report notes that 1.0 mg/L sits below the 1.5 mg/L maximum acceptable concentration yet exceeds the 0.7 mg/L fluoridation target communities aim to apply.
Mistake to avoid: Citing the 1.5 mg/L maximum acceptable concentration as the operational fluoridation target for programs.
Reference: [11]
23. Fluoride assessment includes multiple exposure sources
Fluoride exposure assessment considers water together with food, beverages and dental products; the maximum acceptable concentration is protective when recommendations on other sources, such as dental products, are followed, with children aged 1-4 years the focus of fluorosis risk.
Worked example: An assessor documenting a young community's fluoride exposure records the municipal water concentration and notes that young children may swallow toothpaste, rather than treating the water level alone as total exposure.
Mistake to avoid: Reporting only the drinking-water concentration as a young child's complete fluoride exposure assessment.
Reference: [11]
24. Match fluoride monitoring to the water context
Municipalities that fluoridate should establish a full monitoring and reporting program to confirm the target level is applied at all times, whereas naturally occurring levels, which are not expected to change rapidly, generally require less frequent monitoring; elevated natural fluoride warrants a plan emphasizing young children.
Worked example: A fluoridating city builds routine monitoring and reporting into its operations even though its newest sample meets the 0.7 mg/L target, unlike a neighbouring community with stable naturally occurring fluoride levels.
Mistake to avoid: Applying the relaxed monitoring expectation for stable natural fluoride to an actively fluoridated supply.
Reference: [11]
25. Budget impact answers affordability rather than value
Budget impact analysis estimates whether an intervention is affordable to a constrained budget, while cost-effectiveness analysis is one economic approach relating monetary cost to additional health benefit. The source guidance is pharmaceutical; the dental example only illustrates the general affordability-versus-value distinction, not a dental method.
Worked example: A hypothetical fluoride varnish program adds $20,000 to a health unit's annual dental budget; this shows a budget consequence, while its value would be examined through an economic analysis such as cost-effectiveness analysis.
Mistake to avoid: Treating a small projected budget increase as proof that a program is cost-effective.
Reference: [14]
26. The budget holder's perspective determines included costs
A budget analysis must state whose budget it represents and apply that cost scope consistently. The source's exclusions concern pharmaceutical drug-plan submissions; a dental illustration may define a program perspective but must not mix in costs outside it.
Worked example: A hypothetical school dental screening budget, defined from the program's perspective, records $12,000 in program spending and lists $3,000 of family travel costs separately, without adding them into the program total.
Mistake to avoid: Adding household travel costs into a program-perspective total without stating or changing the perspective.
Reference: [14]
27. Compare financial scenarios and disclose assumptions
Estimate a budget change by comparing a defined reference scenario with a proposed scenario over the same period, stating all assumptions explicitly. Document population, time period, uptake and cost assumptions, and justify inputs for alternative scenarios.
Worked example: A hypothetical mobile dental clinic expansion projects $40,000 in annual service costs without expansion and $55,000 with it, so the incremental budget impact is $15,000, with assumed patient uptake stated and alternative uptake scenarios tested.
Mistake to avoid: Reporting the expanded scenario's full $55,000 cost as the incremental budget impact.
Reference: [14]
Planning oral health programs
28. Define participation and decision-making authority
Pre-planning should clarify who participates in planning, why, when and in what way, and determine how decisions are made, by whom, including the role of power and privilege, involving people affected by and able to affect the program.
Worked example: A hypothetical oral health planning group assigns community representatives a defined, resourced role in ranking priorities before any activities are selected, with the group documenting how disagreements between members will be decided.
Mistake to avoid: Describing a final presentation to residents as meaningful involvement when they had no decision-making role.
Reference: [6]
29. Revise plans when the situation changes
Program planning is iterative rather than strictly linear. When new information emerges during planning, earlier steps may need to be revisited and existing planning documents updated, keeping changes connected to the program's goals and assessment evidence.
Worked example: During planning for an oral health screening program, community feedback shows the chosen hall lacks wheelchair access, so the team revises its audience access assumptions, selects an accessible venue, and updates the resource plan and documented responsibilities.
Mistake to avoid: Keeping the original activities unchanged simply because an earlier planning step was already marked complete.
Reference: [6]
30. Choose the audience from assessment findings
The audience is the specific group the program intends to reach, and identifying it should be rooted in the results of the situational assessment, which identifies those most impacted by the situation, not in recruitment convenience.
Worked example: A situational assessment identifies poor access to dental care among homebound adults, so the plan names homebound adults as the audience rather than defaulting to readily available school pupils who were not the assessed group.
Mistake to avoid: Choosing the easiest group to recruit without explaining its relationship to the assessed need.
Reference: [6]
31. Translate broad goals into measurable objectives
A goal reflects the broadest level of results to be achieved, while a well-written objective describes the audience, the outcome, what will be measured, and when, following SMART-C criteria: specific, measurable, actionable, realistic, time-specific, and challenging.
Worked example: The broad goal of improving referral completion becomes the objective of increasing completed referrals among enrolled adults from the measured baseline of 40% to 50% by December, naming the audience, outcome, measure, and timeframe.
Mistake to avoid: Treating an aspiration without a measure or timeframe as an evaluable objective.
Reference: [6]
32. Strategies explain the approach; activities specify actions
Strategies are the approaches used to plan or facilitate change and accomplish the program's goal, while activities are the proposed events or actions undertaken to produce the desired outcomes. Activity choice should fit the intended audience and available resources.
Worked example: For a program goal of improving local oral health, strengthening community action is the selected strategy, and convening resident-led planning meetings is the activity implementing it, chosen because residents are the intended audience and meetings fit available resources.
Mistake to avoid: Listing meetings as activities without explaining how they advance the chosen strategy and goal.
Reference: [6]
33. Check whether inputs can support delivery
Inputs are the quantifiable resources, or ingredients, invested in a program, including staffing, materials and facilities. Planners must verify that sufficient resources exist to implement the planned activities, reconciling calculated requirements against availability; sufficiency alone does not establish effectiveness.
Worked example: A varnish clinic plans ten sessions, each needing one hygienist for four hours, so 40 hygienist-hours are required; 32 scheduled hours leave an eight-hour shortfall, counting delivery only, with preparation and travel estimated separately.
Mistake to avoid: Approving a session timetable without reconciling its calculated staffing requirement against available hours.
Reference: [6]
34. Outputs count delivery; outcomes describe change
Outputs are the immediate, measurable results of a program's activities, quantified numerically or as percentages. Outcomes are measurable positive or negative changes experienced by the audience, which may be short-term, intermediate or long-term, and require appropriate evaluation before attribution to the program.
Worked example: A program records that 75 caregivers attended brushing demonstrations, an output; a measured improvement in caregivers' demonstrated brushing technique is an outcome, still requiring appropriate evaluation before being attributed to the program.
Mistake to avoid: Claiming children's oral health improved because workshop attendance numbers were high.
Reference: [6]
35. Select indicators that measure the objective
Indicators are specific, observable and measurable characteristics demonstrating achievement of objectives; they should be important, obtainable, reliable and valid, and must measure the intended objective. Proportions and rates need defined numerators and denominators, while simple counts do not.
Worked example: For an objective on reducing adolescents' daily sugary-drink intake, use the proportion of surveyed students reporting daily intake, with defined numerator, denominator, data source and observation period, rather than counting posters displayed.
Mistake to avoid: Selecting leaflet distribution counts as the indicator for a referral-completion objective.
Reference: [6]
36. Use a logic model to test connections
A logic model graphically depicts relationships among program components—goals, objectives, populations, strategies, activities and indicators—and verifies the program's logic, including whether activities advance strategies and indicators demonstrate objectives. It may include assumptions and external factors, but coherence does not establish causal effectiveness.
Worked example: A school sealant program's logic model links trained staff, sealant application clinics and retention checks to measured sealant retention; the connections are logically coherent, but this coherence alone does not establish that sealants cause the measured outcomes.
Mistake to avoid: Treating a coherent, well-drawn logic model as proof of the program's causal effectiveness.
Reference: [6]
Program management, evaluation and quality assurance
37. Make operational responsibilities explicit
An implementable program plan specifies which actions will occur, who is responsible for each, when they happen, what resources are required, and how the program will be communicated. Assignments must fit available staffing capacity and the current program context; this concerns operational accountability, not employment law or professional scope of practice.
Worked example: A coordinator confirms the community hall by November 3, an educator delivers the first brushing session on November 10, and the supervisor approves printed materials by October 27.
Mistake to avoid: Writing that the team will arrange school fluoride-education sessions without naming a responsible person or completion date.
Reference: [6]
38. Diagnose implementation before judging results
Process evaluation documents and assesses a program's reach, audience recruitment and retention, perceived quality, acceptability, and fidelity of implementation. Delivery figures are process measures, not effectiveness estimates, and implementation findings inform interpretation without by themselves resolving causal questions about outcomes.
Worked example: Eight of ten planned brushing sessions occurred, so delivery fidelity was 80%; attendance records then show whether the intended preschool audience was actually reached.
Mistake to avoid: Concluding a fully delivered program was ineffective because outcomes stayed unchanged, without ever checking whether implementation actually occurred as planned.
Reference: [6]
39. Focus evaluation questions on purpose and use
Evaluation questions should be limited to those answerable in a timely manner with available resources and appropriately rigorous methods, aligned with the evaluation's purpose and intended uses. Unlike research, which primarily seeks generalizable knowledge, evaluation aims to improve programs and support decision-making, while still not concealing important limitations or affected groups' information needs.
Worked example: Before commissioning a long-term impact study, a new outreach program first asks whether intended residents were reached and why participation was low.
Mistake to avoid: Collecting extensive resident survey data without first identifying which decisions the evaluation findings are supposed to inform.
Reference: [7]
40. Choose evaluation designs for the question
Experimental designs use random assignment; quasi-experimental designs compare nonequivalent groups or time points without randomization; observational designs typically address noncausal questions such as implementation quality. No design is better under all circumstances: match the design to purpose, questions, budget, timeline, setting, and ethical context, remembering each category retains its own validity limitations.
Worked example: A dental sealant evaluation compares children who received sealants with a nonrandom waiting-list group, making the design quasi-experimental rather than a randomized experiment.
Mistake to avoid: Labeling a comparison of participants with a waiting-list group as a randomized experiment without checking how participants were actually assigned.
Reference: [7]
41. Build credible evidence with understandable instruments
Collection methods should connect directly to the evaluation questions, and instruments should be reviewed with people familiar with respondents so wording is understood. Such review improves suitability but does not itself establish validity; data risks, confidentiality, handling arrangements and applicable ownership requirements must also be explained and respected.
Worked example: Before surveying caregivers about preventive dental visits, the team reviews draft items with community representatives, who note that 'regular check-up' is unclear about frequency, so the item is reworded to specify intervals.
Mistake to avoid: Assuming a technically worded questionnaire automatically measures the intended construct accurately for all respondents.
Reference: [7]
42. Validate calculations and model inputs
Quality assurance for a budget model includes checking equations and parameters against their sources and testing that calculations are implemented and performed correctly. Correct arithmetic alone does not validate unsupported population, uptake or unit-cost assumptions.
Worked example: A program analyst validating a sealant budget recalculates 180 planned visits at $60 each, confirming the displayed $10,800 total matches a hand-checked product before the model is used.
Mistake to avoid: Trusting a polished spreadsheet's outputs without checking its formulas against documented sources.
Reference: [14]
Evidence synthesis and useful communication
43. Synthesize findings within the wider evidence
Knowledge synthesis contextualizes and integrates individual studies within the larger body of knowledge on a topic. Differences in populations, interventions and contexts must be retained when judging applicability; this concept describes synthesis's purpose, not a full systematic review method.
Worked example: A favourable local school-based fluoride education evaluation is interpreted alongside other published studies and its specific setting, rather than being reported as a stand-alone result applicable everywhere.
Mistake to avoid: Selecting one convenient favourable finding while ignoring the broader evidence relevant to the decision.
Reference: [10]
44. Knowledge exchange requires mutual learning
Knowledge exchange involves interactions between knowledge users and researchers that result in mutual learning, with both sides contributing. Community insight adds contextual understanding without automatically proving causation, and the exchange's form and intensity should fit the research and users' needs.
Worked example: During a fluoridation review, researchers share survey results while water-utility staff explain local supply constraints; both sides adjust their understanding, and the revised interpretation reflects contributions from each group.
Mistake to avoid: Labelling a one-way presentation of findings 'exchange' when users cannot contribute.
Reference: [10]
45. Identify users by the decisions they can make
CIHR defines a knowledge user as an individual likely to use research results to make informed decisions about health policies, programs or practices; relevance depends on the decision and information need, not merely organizational seniority.
Worked example: A dental public health service manager who can revise clinic booking arrangements is a relevant user of appointment-access research, alongside community representatives affected by access barriers.
Mistake to avoid: Defining the intended audience solely as researchers who might cite the published work.
Reference: [10]
46. Distinguish integrated and end-of-grant translation
Integrated knowledge translation involves knowledge users as research-team members across many stages of the process; end-of-grant translation plans how completed findings will reach users. Even iKT programs still require a dissemination plan.
Worked example: Community partners help frame research questions and interpret results, and the team still prepares briefings and plain-language summaries so other relevant audiences receive the completed findings.
Mistake to avoid: Assuming early user collaboration removes the need for any dissemination plan.
Reference: [10]
47. Set a translation goal beyond producing a document
A knowledge translation goal should be clear, concrete and appropriate to the potential findings and target knowledge-user audiences; producing a report is a delivery activity, not the intended change in awareness, behaviour, policy or practice.
Worked example: For educators, the goal may be improved awareness of referral barriers, with a stakeholder briefing serving only as the delivery method for that awareness goal.
Mistake to avoid: Equating publication of a report with demonstrated knowledge use or achieved change.
Reference: [10]
48. Tailor the message and medium to the audience
Dissemination means identifying the appropriate audience for research findings and tailoring both message and medium to that audience, considering knowledge needs, context and preferences, without altering the findings or hiding uncertainty.
Worked example: The same access findings become a concise options briefing for managers and an accessible plain-language explanation for residents, with the underlying evidence unchanged.
Mistake to avoid: Sending the same technical report to every audience and assuming it is usable.
Reference: [10]
49. Match action messages to evidence strength
Research moves into practice only when evidence strength is sufficient; the guide sets no universal threshold. A preliminary local study may identify a question for further investigation rather than support a universal service change, and application must also fit ethical, social, and regulatory conditions.
Worked example: A resident reports that a single-clinic pilot suggested fewer missed appointments with text reminders and frames it only as a question for larger study, since application must fit applicable ethical and regulatory conditions.
Mistake to avoid: Recommending universal adoption of a reminder program from one preliminary clinic pilot because administrators want quick action.
Reference: [10]
50. Explain rates without losing their units
A person-time rate describes how quickly new disease occurs, so plain-language explanations must keep the numerator, denominator, and time unit. A rate of 20 per 1,000 person-years is not automatically a 2% one-year individual probability; tailor wording to the audience while retaining the exact measure.
Worked example: A hygienist explains to a community board that four new caries cases arose during 200 person-years of follow-up, meaning 20 new cases per 1,000 person-years, describing the observed pace rather than any individual's yearly risk.
Mistake to avoid: Telling councillors the rate is simply '2% risk' and dropping the person-years time dimension entirely.
51. Report methods and limitations with the finding
Evaluation reporting should share methods, findings, recommendations, strengths, and limitations together, so users can judge how far conclusions travel. Conclusions stay bound to the design and context, and limitations should clarify interpretation rather than excuse unsupported recommendations.
Worked example: An access briefing states that its findings came only from clinics volunteering to participate, that non-participating settings were not assessed, and that recommendations therefore apply cautiously until broader data confirm them.
Mistake to avoid: Announcing that mobile clinics solved rural access gaps without describing the survey methods or their limits.
Reference: [7]
52. Communicate the full pattern of results
Findings should be reported holistically, within the evaluation's agreed questions and context, instead of selecting results that favour a preferred position. Balanced reporting still means unsupported interpretations do not receive equal weight with supported conclusions.
Worked example: A school sealant evaluation finds higher appointment completion but notably heavier staff workload; the options briefing to the health authority reports both outcomes together within the questions the evaluation was designed to answer.
Mistake to avoid: Presenting only the improved completion figures to win program expansion while omitting the documented workload burden.
Reference: [7]
53. Use timely feedback throughout evaluation
Dissemination is a cycle conducted regularly, not a final act. Throughout an evaluation, evaluators can share in-process data, hold user check-ins and run feedback sessions, labelling interim analyses and provisional interpretations accurately so learning is supported without overstating unfinished results or compromising confidentiality.
Worked example: During a community fluoride varnish program's implementation, the evaluator shares a provisional attendance pattern, clearly stating its uncertainty, so the team can investigate recruitment problems while data collection continues.
Mistake to avoid: Delaying every useful finding until the program can no longer respond to it.
Reference: [7]
54. Resource the communication and education plan
Knowledge translation plans must be deliverable: the team needs expertise appropriate to the actual delivery method and audience, plus financial, human and access resources. Feasibility planning shows activities can be accomplished; it does not claim any particular education format guarantees learning.
Worked example: A multilingual oral health workforce education session is budgeted with interpreter support, an experienced facilitator and adequate delivery time, rather than assuming a prepared slide file alone makes the session deliverable.
Mistake to avoid: Planning an elaborate learning product without the personnel or budget to deliver it.
Reference: [10]
Education and ethical community research
55. Protect voluntary participation and withdrawal
Within TCPS 2 research ethics requirements and applicable law, consent must be voluntary and withdrawable at any time, and pre-existing entitlements to care, education and services must not be prejudiced. Participants may request withdrawal of their data or materials, though not every completed action can be reversed, and dependency or power relationships must be judged from the prospective participant's perspective.
Worked example: A dental clinic invites patients to a screening survey, stating clearly that care is independent of participation; a patient who declines continues receiving the same care without any comment or consequence.
Mistake to avoid: Implying that joining or remaining in the study is required for services.
Reference: [8]
56. Maintain consent as information and capacity change
Consent is an ongoing process: researchers must provide all information relevant to continuing participation, including changes to risks or benefits, respecting applicable legal requirements and the approved consent process. Capacity-based rather than age-based consent means developing, diminishing or fluctuating capacity can require changes in how continuing consent is obtained.
Worked example: A longitudinal oral health study changes its data-sharing arrangements mid-project; researchers explain the change to participants so they can reconsider whether to continue, rather than relying on the original signature.
Mistake to avoid: Treating an initial signature as permission for every later change to the study.
Reference: [8]
57. Handle material incidental findings within consent
An incidental finding is a discovery made during research that falls outside the study's objectives. It is material when reasonably determined to carry significant welfare implications, and researchers shall disclose material findings to the participant within the limits of that participant's consent.
Worked example: During a community oral health survey, a researcher notices an unusual lesion on a participant's photograph; lacking interpretive expertise, the team consults an appropriate specialist before deciding what to disclose within consent limits.
Mistake to avoid: Dismissing an unexpected lesion as irrelevant simply because it was not a planned study outcome.
Reference: [8]
58. Community engagement does not replace individual consent
For research on lands under First Nations, Inuit or Métis authority, researchers shall seek leaders' engagement except as provided under Articles 9.5–9.7. Institutional REB review, subject to Article 8.1, and any responsible community review must precede recruitment; these collective steps accompany, never replace, each individual's consent.
Worked example: After institutional review, subject to Article 8.1, and community review for a planned adult oral health survey on First Nations lands, one invited adult declines, so that person is not enrolled despite collective approvals.
Mistake to avoid: Treating a band council's collective approval as though every adult resident had personally consented to participate.
Reference: [9]
59. Agree on research responsibilities and knowledge governance
When a community formally engages a research team through a designated representative, both parties' terms and undertakings should be set out in a research agreement before recruitment, including data management responsibilities and any restrictions on access to or use of traditional or sacred knowledge.
Worked example: Before recruiting Elders for an oral knowledge project, the researcher and the community's designated representative sign an agreement specifying who may access interview recordings, how restricted teachings are handled, and how results return.
Mistake to avoid: Publishing sacred knowledge online without checking the agreement's restrictions on disclosure and access.
Reference: [9]
60. Build workforce evaluation skills through participation
Interest holders who take part in planning and implementing an evaluation can build evaluation knowledge and skills. Because participation can support capacity building without guaranteeing competence, learning activities should fit participants' roles and the evaluation's questions and context.
Worked example: Public health dental hygienists co-design referral-rate indicators with evaluators and help interpret pilot findings, gaining practical experience in using evaluation evidence while contributing their operational knowledge of clinic workflows.
Mistake to avoid: Counting attendance at a results briefing as proof that staff can apply evaluation skills.
Reference: [7]
Sources
- RCDC
- NDSE Resources
- NDEB
- CDC Principles of Epidemiology
- CDC Principles of Epidemiology
- Public Health Ontario
- CDC Program Evaluation Framework, 2024
- TCPS 2 (2022), Chapter 3
- TCPS 2 (2022), Chapter 9
- CIHR
- Health Canada
- Canadian Dental Care Plan
- Health Canada
- Government of Canada