Use these selected concepts to connect population assessment, prevention, health systems, ethics, research, communication, and policy in your preparation for the American Board of Dental Public Health (ABDPH) Diplomate Examination pathway. Work through the foundations before applying them to program and policy decisions. The examples illustrate how to interpret information and justify decisions within the limits of the evidence.
Population health foundations and determinants
1. The community as the unit of dental public health
Dental public health treats a population as its patient, planning prevention, education, research, and program administration through organized community effort rather than one-to-one care. Community clinic services can be a deliberately planned part of that work, yet clinic treatment by itself covers only a slice of the specialty's broader population scope.
Worked example: A county health department reviews clinic utilization data and adds a school screening survey, revealing children with untreated decay who had never visited the clinic; the assessment therefore plans for the whole county population rather than only attending patients.
Mistake to avoid: Treating clinic attendance records as a complete account of community oral health needs, ignoring residents who never seek or reach care.
Reference: [4]
2. Needs assessment includes resources and unmet need
Planning requires more than counting disease. A complete assessment describes current and potential problems, inventories the services and resources a community already has, and identifies where needs go unmet or where capacity sits unused because of access barriers or delivery-system shortcomings.
Worked example: Surveyors document a high prevalence of untreated decay, then inventory the county's clinics, Medicaid participation, and referral logs, learning that referred children frequently never receive appointments; the assessment reports the disease burden together with underused capacity and broken referral pathways.
Mistake to avoid: Recommending additional clinic capacity before checking whether existing services are inaccessible, underused, or failing to complete referrals.
Reference: [8]
3. Separate individual barriers from system barriers
Access analysis should distinguish family-level barriers, such as knowledge or attitudes, from system barriers, including financing, insurance coverage, provider distribution, transportation, and other social determinants of health. Separating these levels matters because the appropriate response depends on the identified obstacle; the presence and importance of each barrier require assessment in the relevant community before any intervention is proposed.
Worked example: A program director learns that families understand referral instructions but no nearby Medicaid-participating dentist is accepting new children; she investigates transportation options and provider participation rather than repeating an oral hygiene education campaign aimed at parental knowledge.
Mistake to avoid: Attributing every missed appointment to parental attitudes without assessing insurance acceptance, provider availability, and transportation barriers in the community.
Reference: [9]
4. Health literacy can affect access before treatment
Health literacy can influence a person's ability to locate health care providers and services, fill out health forms, and share personal health information with providers, so administrative and communication demands can create access barriers before any clinical encounter occurs. These demands should be assessed when investigating access, and an individual's literacy should never be inferred from demographic membership or education alone.
Worked example: A family reaches a clinic easily but cannot complete its dense enrollment form, so the coordinator revises and user-tests the form with plain language; simplifying this pre-treatment step addresses an access obstacle that appointment availability alone never explained.
Mistake to avoid: Counting available appointments as proof that the enrollment process is usable for families with limited health literacy.
Reference: [16]
5. Equity requires examining participation and power
An equitable evaluation examines who participates, whose perspectives are missing, and what historical or current power differences may affect the process. Participation should inform decisions rather than serve only as an attendance count, and a technically diverse committee does not necessarily represent the communities the program is intended to benefit.
Worked example: A state fluoride varnish program evaluation invites only program managers and funders to interpret findings; adding caregivers of children served and reviewing how priorities were set addresses the missing perspective and possible power imbalance.
Mistake to avoid: Treating attendance counts as sufficient evidence of equitable participation without examining whose perspectives are missing or how contributions are considered.
Reference: [7]
Ethics and protection of participants
6. Distinguish research from individual clinical practice
Research is designed to test hypotheses and contribute to generalizable knowledge, while practice aims to enhance the well-being of particular individuals. When an activity combines research with clinical practice, it should undergo review for the protection of human subjects, because the clinical setting does not remove the research element or its review requirement.
Worked example: A dental clinic adds a structured protocol comparing two sealant schedules to test which reduces caries more, intending publication; because the activity contains a research element, it should undergo human subjects review rather than proceed as routine care.
Mistake to avoid: Labeling a hypothesis-testing protocol as ordinary care solely because every enrolled participant continues to receive dental services.
Reference: [13]
7. Respect autonomy while protecting diminished autonomy
Respect for persons requires acknowledging autonomous choices and providing appropriate protection when autonomy is diminished. The extent of protection should depend on the risk of harm and the likelihood of benefit, and judgments about autonomy are situation-dependent, should be periodically reevaluated, and cannot be resolved by group membership alone.
Worked example: A researcher plans interviews with older adults whose capacity may vary day to day; rather than excluding the whole group by label, the team assesses each person's understanding at the time and calibrates safeguards to the study's risk and potential benefit.
Mistake to avoid: Assuming a diagnosis automatically removes autonomy, so consent can be skipped without any individualized capacity assessment.
Reference: [13]
8. Consent requires information, comprehension, and voluntariness
Informed consent is a process involving adequate information, comprehension, and voluntary choice, including the opportunity to ask questions and withdraw at any time. A signature alone does not establish all three elements, and this standard concerns research consent rather than serving as a universal template for every clinical or surveillance activity.
Worked example: In a survey of dental clinic patients, a staff member hands over a dense form, rushes the explanation, and implies that refusing could affect future appointments; even with a signed form, comprehension and voluntariness remain unresolved.
Mistake to avoid: Treating a signed consent form as proof that the participant understood the study and volunteered freely.
Reference: [13]
9. Assess both the probability and magnitude of harm
Risk/benefit review weighs both the probability that a harm might occur and how serious it would be, across psychological, physical, legal, social, and economic dimensions. Anticipated societal benefit never removes the duty to protect participants or minimize unnecessary risks, so a study without clinical procedures can still pose real, assessable harm.
Worked example: An interview study about workplace dental benefits involves no physical procedures, but participants' disclosures could jeopardize their employment or community standing. Reviewers therefore assess the likelihood and severity of these social and economic harms and require confidentiality safeguards rather than classifying the protocol as risk-free.
Mistake to avoid: Calling a survey study harmless because it involves no clinical procedures, while overlooking possible social and economic harms to participants.
Reference: [13]
10. Participant selection must be fair
Justice requires fair procedures and outcomes in selecting research subjects, meaning recruitment should follow from the scientific question and an equitable sharing of burdens and benefits. Vulnerable populations may participate when the choice is substantively justified and proper protections exist, but convenience alone is insufficient, and fair treatment after enrollment cannot repair an unfair recruitment strategy.
Worked example: A state oral health survey team proposes recruiting residents of one group home because staff access is easy, yet the research question concerns all low-income older adults. The team broadens recruitment to multiple community settings and adds consent safeguards before seeking approval.
Mistake to avoid: Recruiting a dependent institutional population only because access is convenient, then assuming fair treatment after enrollment cures the unfair selection.
Reference: [13]
11. Confidentiality and data-sharing governance are separate duties
Sharing surveillance data among public health partners is encouraged, but only in ways consistent with law and with protection of personal confidentiality and privacy. Data ownership must be clear and explicit data-use agreements established; however, an agreement alone does not create legal permission for every disclosure, since applicable agency rules still govern, and a public health purpose is not unlimited authority.
Worked example: Two state agencies draft a data-use agreement to share identifiable oral cancer screening records, but the disclosure still requires review under each agency's confidentiality rules. The team confirms legal authority, clarifies data ownership, and documents permitted uses before any records are transferred.
Mistake to avoid: Treating a signed data-use agreement as automatic legal permission to share identifiable records without checking applicable agency confidentiality rules.
Reference: [8]
Surveillance and representative data
12. Surveillance produces information for action
Surveillance is an ongoing cycle: health data are systematically collected, analyzed, and interpreted, then communicated to those positioned to act. Collection alone leaves the cycle incomplete; the stated purpose and intended audiences determine which analyses and reports are needed, so record volume is not success.
Worked example: A county stores annual school screening records but never analyzes or reports them. Staff add trend analysis and a yearly summary for program planners and school administrators, completing the interpretation and dissemination functions surveillance requires.
Mistake to avoid: Judging a surveillance system successful by the number of records collected while ignoring whether findings ever reach decision-makers.
Reference: [8]
13. A survey snapshot becomes surveillance through repeated use
A single cross-sectional survey records the oral health of a defined population at one point in time, so it cannot by itself establish change. Surveillance emerges when such surveys are repeated on separate samples and aggregated with other data sources, allowing population-level monitoring of trends while remaining distinct from individual follow-up of the same people.
Worked example: A state health department repeats its Basic Screening Survey of randomly sampled third-graders every five years and combines each round with Medicaid dental-visit data and water fluoridation records, producing population-level trend estimates without tracking any individual child across rounds.
Mistake to avoid: Presenting two cross-sectional surveys of different samples as proof that the same children's decay experience changed over time.
Reference: [8]
14. A surveillance plan connects purpose to operations
A written surveillance plan functions as an operational roadmap, not merely a wish list of indicators. It should specify the system's purpose, objectives, indicators, data sources, target populations, operating resources, collection schedule and protocol, analysis methods, data use and dissemination, privacy and confidentiality practices, and an evaluation protocol, so staff can actually build, maintain, and assess the system.
Worked example: A county drafts a plan naming its purpose, the untreated-decay indicator, the school-based screening source, a biennial collection schedule, the analyst time required, a public report each cycle, de-identification procedures, and a five-year evaluation, giving staff a workable roadmap.
Mistake to avoid: Submitting an indicator list as the surveillance plan while omitting data sources, schedules, resources, privacy practices, and evaluation.
Reference: [8]
15. Identify data gaps before collecting new data
Developing surveillance starts with an inventory of existing oral health data sources and a review of indicators already reported for the U.S. population. Only after selecting indicators and identifying where data are unavailable or unanalyzed should a program prioritize new collection, since suitable secondary data may already exist but simply not have been used.
Worked example: Before drafting a new adult questionnaire, a program inventories BRFSS records and finds state-level dental-visit data already collected every two years; it redirects its limited field resources toward screening older adults in long-term care, an indicator no existing source covers.
Mistake to avoid: Commissioning a primary survey for an indicator that an existing state dataset already measures but the program never examined.
Reference: [8]
16. Match data quality and timeliness to the decision
Because no data are perfect and perfecting them is costly, surveillance data need only the quality, resolution, and timeliness that the stated public health objective requires. Systems should therefore match data standards to intended use, since delaying action for refinements irrelevant to the decision can undermine actionable monitoring rather than improve it.
Worked example: A coalition deciding whether to fund a school sealant program needs county-level untreated-decay estimates within one budget cycle; using the available screening survey now, while documenting its sampling limitations, serves that decision better than waiting years for a more precise study.
Mistake to avoid: Automatically selecting the most detailed dataset for a routine annual report despite its multi-year reporting delay making it untimely.
Reference: [8]
17. NHANES weights represent unequal selection
A base sampling weight generally equals the reciprocal of a sampled person's probability of selection, but NHANES final weights go further: they adjust for the complex multistage design, including oversampling, for interview or examination nonresponse, and for post-stratification to Census Bureau population totals. Weighted estimates therefore represent the U.S. civilian noninstitutionalized population, and a reciprocal calculation alone is not a final NHANES weight.
Worked example: Example: A participant with a hypothetical selection probability of 0.005 starts with a base weight of 200, but subsequent nonresponse and post-stratification adjustments may shift the final weight, so the analyst should use the released weight variable rather than recomputing reciprocals.
Mistake to avoid: Avoid: Assuming that interview or MEC examination nonresponse adjustments also correct every unanswered item; only designated subsample weights add component-nonresponse adjustments.
Reference: [6]
18. Select NHANES weights for the smallest relevant component
Choose the sample weight tied to the variable collected on the smallest number of respondents, the analysis's least common denominator, because that weight carries the adjustments for that group's selection and nonresponse. Component subsamples have designated weights, and some questionnaires administered during the MEC require MEC exam weights, so confirm recommendations in each component's documentation file, since weight variables vary by survey cycle.
Worked example: Example: An analysis links in-home interview demographics with fasting triglycerides from a randomly selected morning subsample; the designated fasting subsample weight, not the interview weight, applies because the laboratory variable was collected on the smaller respondent group.
Mistake to avoid: Avoid: Selecting a weight solely because it is already present in the demographic file, since subsample weights reside on their component data files.
Reference: [6]
19. Pool survey cycles only under the appropriate rules
For comparable ordinary two-year cycles from 2001–2002 onward, analysts generally construct multi-year weights by dividing each two-year weight by the number of cycles combined, after verifying item comparability and assuming no trend. Special cases differ: analyses including 1999–2000 must begin with NCHS's provided four-year weights, the 2017–March 2020 pre-pandemic files cover 3.2 years, and combining August 2021–August 2023 with earlier cycles is generally not recommended.
Worked example: Example: Pooling three compatible ordinary two-year cycles, an analyst divides each hypothetical two-year weight of 6,000 by three, producing a 2,000 multi-year weight, after confirming the variables' wording, methods, and eligibility criteria are comparable across all three cycles.
Mistake to avoid: Avoid: Applying the divide-by-cycle-count rule to every release without checking duration, comparability, and trends, especially for pre-pandemic or gapped cycles.
Reference: [6]
Evidence appraisal and research design
20. Interpret surveillance patterns in context
Surveillance findings gain meaning only when read against historical trends and current oral disease epidemiology. A current observation must be compared with prior comparable data and present disease context; this comparison shows whether a pattern is emerging or persistent, without establishing cause.
Worked example: A district's untreated-decay finding initially looks like a new problem, but earlier comparable reports show the same elevated burden repeatedly, so staff interpret it as a persistent problem rather than a newly emerging increase.
Mistake to avoid: Calling a current finding a new trend without examining the historical record and current disease context.
21. Check evidence quality and applicability separately
Critical appraisal examines two separate questions: the quality of the evidence, judged by recognizing study-design limitations and factors influencing validity, and its applicability to the specific research question. Topical relevance never resolves a study's design limitations, and validity judgments must address the actual question asked.
Worked example: A county asks about dental access among older adults, but a report studies schoolchildren. The reviewer logs the population mismatch under applicability and appraises design limitations separately; dental relevance does not answer the county's question.
Mistake to avoid: Treating a study as suitable merely because it discusses the same broad oral health topic.
Reference: [4]
22. Specify measures of population oral health
Population oral health research should incorporate epidemiological principles into its design and specify indices depicting population oral health status as outcome measures. Measures are chosen for the stated research question; no particular index or scoring procedure is mandated, so the outcome must fit the question.
Worked example: A proposal asks whether a program changes population oral health status but records only appointment attendance. The team adds a specified oral health status index as the outcome, keeping attendance only as service-use information.
Mistake to avoid: Presenting attendance alone as the measured oral health status outcome when the question concerns population disease status.
Reference: [4]
23. Compare new interventions with current practice
Appraising a new public health intervention covers two steps: assessing how the option performs, then comparing those findings with current practice. Evidence describing the new option alone leaves the decision unfinished, and comparisons should carry the original populations, settings, and limitations.
Worked example: A program weighs replacing its outreach approach. The only report on the candidate describes promising results without any comparison to the current approach, so the team records the missing comparison and withholds any improvement claim.
Mistake to avoid: Recommending replacement because an option is new or shows favorable results, without comparing findings with current practice.
Reference: [4]
24. Review findings depend on study selection
Appraising a systematic review includes examining its search period, eligibility criteria, and sequential screening stages rather than judging credibility by the initial number of records retrieved. Conclusions rest on the included studies and the selection process, not on every record the search identified.
Worked example: A hypothetical review of school-based fluoride varnish programs identifies 90 records, excludes most at title screening, and includes seven eligible studies. Its conclusions rest on those seven studies and the documented selection process, not on 90 independent results.
Mistake to avoid: Treating every retrieved record as evidence supporting the review's conclusion, ignoring exclusions made during screening stages.
Reference: [10]
Program planning and evaluation
25. Evaluation and research differ in primary purpose
What separates evaluation from research is where findings are meant to go, not the techniques used. Data gathered to guide a program's own next decisions serve evaluation; data gathered to answer questions that hold beyond one program serve research; one method can serve either aim.
Worked example: A community dental clinic interviews patients about why sealant referrals stall, intending to revise its own booking process; that improvement purpose makes the activity evaluative, even though interviews also appear in research.
Mistake to avoid: Labeling an activity research solely because it uses statistical tests, without examining the findings' intended use.
Reference: [7]
26. A logic model connects activities with intended outcomes
A program description and logic model lay out needs, inputs, activities, contextual factors, stage of development, and outcomes across short-term, intermediate, and long-term horizons. This narrative and roadmap make the program's assumed causal pathway explicit and assessable, but the diagram itself documents intent only; it is not evidence that the intended changes actually occurred.
Worked example: For a school-based varnish program, hygienist time and donated supplies are inputs, classroom visits are activities, and teeth receiving varnish are service outputs; improved caregiver knowledge is an intended short-term change, with reduced untreated decay remaining a proposed outcome to assess.
Mistake to avoid: Avoid: Treating the number of services delivered, an output, as proof that participants' health, knowledge, or behavior actually changed.
Reference: [7]
27. Match evaluation purpose to program maturity
Evaluation questions should fit the program's stage of development. Emerging programs that are still adapting benefit from evaluation that supports innovation and development, while examining effectiveness is most appropriate for mature programs with clear, stable implementation. Matching purpose and design to context avoids judging a shifting intervention as though its delivery were fixed.
Worked example: A teledentistry pilot keeps revising its appointment workflow each month, so evaluators document adaptations to support development; an effectiveness examination is most appropriate for a mature program whose implementation has become clear and stable.
Mistake to avoid: Avoid: Designing an evaluation around stable delivery assumptions when the program's implementation is continually changing.
Reference: [7]
28. Evidence collection must answer the evaluation question
Credible evidence comes from clearly connecting the evaluation questions and purpose statement to the data collection strategy, including methods, indicators, and sources. Reviewing instruments with people familiar with respondent groups and considering culturally appropriate options increases the likelihood that feasible, understandable data will actually answer the questions identified as relevant with interest holders.
Worked example: An evaluation asks whether referred children completed dental visits, but the available dataset records only referrals issued; the team adds clinic completion records so the collected evidence can actually address the stated question about completed referrals.
Mistake to avoid: Substituting an easily counted activity, such as referrals issued, for the outcome actually under evaluation, like completed referrals.
Reference: [7]
29. Specify how findings will be judged
Before results arrive, program partners should agree on and document which expectations will guide interpretation of the findings. When justified expectations cannot be set in advance, the team instead documents the interpretation procedure itself, including who will participate and how discussions will be structured. This transparency makes later judgments about program performance accountable rather than improvised.
Worked example: Before a school sealant program's annual review, its evaluation partners record which coverage objectives will define acceptable performance and agree that any unexpected results will first be discussed jointly before conclusions are drafted.
Mistake to avoid: Inventing a favorable benchmark after seeing the program's results, which makes the later judgment appear more objective than it actually was.
Reference: [7]
30. Interpret results within their context
Conclusions should reflect the full pattern of findings together with the setting in which the evaluation occurred. Selectively highlighting favorable indicators, overemphasizing negative ones, or applying results to questions the evaluation never addressed misrepresents the evidence. Reporting relevant strengths, limitations, and implementation changes keeps the interpretation honest and useful for decision makers.
Worked example: A school fluoride varnish evaluation finds attendance rose but repeat-visit rates fell. Reporting both trends, alongside a staffing change that affected recall scheduling, gives the district a more accurate basis for planning than announcing the attendance gain alone.
Mistake to avoid: Using one favorable indicator to characterize every program outcome, which presents a selective picture rather than the overall findings.
Reference: [7]
31. Plan for findings to be used
Using evaluation findings requires deliberate preparation: tailoring products for intended audiences, sharing interim results with their provisional status, facilitating group interpretation, and revisiting planned actions after implementation. Dissemination is a recurring cycle rather than a single final report, and follow-up helps ensure recommendations are actually understood and applied rather than merely received.
Worked example: Midway through a community oral health screening evaluation, the evaluators share an emerging appointment bottleneck with program staff, clearly labeling the analysis as provisional, then reconvene after the revised referral process has operated to review what changed.
Mistake to avoid: Assuming that distributing a written report establishes that its recommendations were understood, accepted, or applied by the intended users.
Reference: [7]
Population prevention programs
32. Sealants address susceptible chewing surfaces
Dental sealants are an evidence-based preventive intervention for the chewing surfaces of molars, where pits and fissures collect plaque and food particles. Program claims should stay within that anatomical and intervention scope, and clinical assessment and placement belong to qualified oral health professionals working under applicable standards and consent procedures.
Worked example: A school-based sealant program describes its service as preventing decay on molar chewing surfaces, delivered by licensed clinicians under district consent procedures, and it does not suggest that sealants alone protect smooth-surface or interproximal areas of the dentition.
Mistake to avoid: Expanding a surface-specific preventive claim into a universal promise that sealants protect children against decay on every tooth surface.
Reference: [9]
33. Prioritize schools with substantial preventive need
Targeting should rest on local assessment identifying schools with high concentrations of children at elevated decay risk and barriers to preventive care, such as low family income or limited access to oral health services. Prioritizing such schools improves efficiency, cost-effectiveness, and equity; convenience of access alone does not justify site selection.
Worked example: A county program reviews local assessment data showing one school serves many children from families with low incomes who lack a regular dental home, while an easier-to-reach school shows lower documented need; the program prioritizes the higher-need school.
Mistake to avoid: Choosing schools based mainly on scheduling convenience or proximity rather than documented preventive need and access barriers identified through local assessment.
Reference: [9]
34. Quality assurance includes technique and appropriateness
Quality assurance spans two dimensions: technical performance, reflected in sealant retention, and appropriateness, meaning children served are those at high risk for decay. Programs support both through staff training and calibration, current infection-control and clinical standards, and data collection. Favorable retention or high service volume alone does not demonstrate adequate quality.
Worked example: A program reports strong one-year retention rates, yet its data show few children from its priority high-need population received sealants; reviewers judge targeting deficient despite favorable technical results, and the program revises its school selection criteria.
Mistake to avoid: Treating a high number of sealants placed as proof of quality assurance while ignoring whether children at high risk were actually reached.
Reference: [9]
35. Retention monitoring requires planned follow-up
Retention cannot be assumed from placement records. Programs should schedule retention checks and re-screen children within one year of initial placement, replacing lost material and arranging appropriate professional follow-up where indicated. This planned follow-up reflects the ASTDD school-program recommendation and forms part of a program's quality-assurance system.
Worked example: A program schedules retention assessments nine months after placement, re-screens participating children within the recommended one-year interval, and arranges professional re-evaluation for children whose sealants are found partially or fully lost.
Mistake to avoid: Assuming sealants remain intact because placement logs exist, without scheduling any retention check or re-screening within one year.
Reference: [9]
36. A referral system must track follow-through
Programs should identify children with treatment needs, facilitate referrals to available qualified providers, and track follow-through with families under written protocols. Preventive services complement rather than replace needed clinical assessment and treatment; a referral counts as completed only when the child's received care is documented.
Worked example: A program issues thirty restorative referrals; eighteen are documented complete, twelve lack documented completion, so staff follow tracking protocols, contact families, and coordinate with community providers until each child's status is resolved.
Mistake to avoid: Recording a referral letter as completed treatment without verifying that the child actually received the needed restorative care.
Reference: [9]
Dental benefits and systems of care
37. Children's Medicaid dental benefits differ from adult benefits
Federal Medicaid rules require dental coverage for all enrolled children through the EPSDT benefit, while adult dental coverage is left entirely to each state's discretion, with no federal minimum requirement. A reviewer evaluating an adult enrollee's benefits must therefore examine the state's adult package directly, because a state's obligations toward children reveal nothing about what adults may receive.
Worked example: A consultant reviewing a state's Medicaid program describes the EPSDT dental entitlement for children and then separately retrieves the state's adult dental schedule, noting that the state may offer adults only emergency extractions without contradicting its children's obligations.
Mistake to avoid: Assuming that a state's robust EPSDT dental program for children guarantees comparable dental coverage for adult Medicaid enrollees in that state.
Reference: [14]
38. An oral screening does not replace a dental examination
Under children's Medicaid guidance, an oral screening performed during a physician's physical examination is a screening tool only and does not substitute for a dental examination by a dentist. A child still requires a dentist referral according to the state's dental periodicity schedule and at other medically necessary intervals, so documented screenings cannot close the examination requirement.
Worked example: A five-year-old's well-child visit records a normal oral screening. The care coordinator still schedules a dentist referral under the state's dental periodicity schedule rather than treating the screening as the required dental examination.
Mistake to avoid: Recording a physician's mouth inspection in the chart and concluding that the child's Medicaid dental examination requirement has already been satisfied.
Reference: [14]
39. EPSDT medical necessity extends beyond the listed plan
For children entitled to EPSDT, states determine medical necessity, and when a screening identifies a condition requiring treatment, the state must furnish the necessary services even if those services do not appear in the state's Medicaid plan. The absence of a service from a plan's routine list therefore does not by itself resolve coverage for an EPSDT child.
Worked example: A screening finds a condition needing a service absent from the state plan's routine list. The reviewer routes the case through the state's medical-necessity process instead of denying coverage solely because the service is unlisted.
Mistake to avoid: Denying an EPSDT child's needed treatment automatically because the service does not appear on the state Medicaid plan's routine list.
Reference: [14]
40. Dental periodicity is not the general screening schedule
Each state develops a dental periodicity schedule in consultation with recognized dental organizations involved in child health care, and the periodicity schedule for other EPSDT services may not govern dental timing. Referrals to a dentist follow the state's dental schedule and also occur at other intervals when medical necessity indicates, so no universal interval applies across states.
Worked example: A coordinator verifying a child's dental visit interval consults the state's dental periodicity schedule rather than its well-child schedule, and refers the child sooner after a caregiver reports a concerning lesion that warrants medically necessary evaluation.
Mistake to avoid: Applying the state's general EPSDT well-child periodicity schedule to dental visits and ruling out additional dentist referrals between routine intervals.
Reference: [14]
41. CHIP coverage depends on program structure
CHIP dental coverage depends on program structure. Medicaid-expansion CHIP programs must provide EPSDT, while separate CHIP programs meet dental requirements through a qualifying or benchmark package covering prevention, restoration of oral structures to health and function, and emergency conditions. Reviewers should identify each program's structure and verify the applicable package before applying its benefit rules.
Worked example: Two children in the same state have CHIP cards from different structures: one enrolled through Medicaid expansion receives EPSDT dental services, while the other in separate CHIP receives the benchmark package, so the reviewer confirms each structure before judging coverage.
Mistake to avoid: Assuming every CHIP program follows identical EPSDT rules without first identifying whether coverage comes through Medicaid expansion or a separate program.
Reference: [14]
Economic evaluation
42. Net cost includes costs averted
Net cost equals an intervention's implementation costs minus averted medical and productivity costs, calculated against a stated comparator. When a more effective intervention has negative net cost, results are reported as net cost savings rather than as a cost-effectiveness ratio. A negative figure alone does not establish improved health; the included cost categories must be stated.
Worked example: A hypothetical school sealant program costs $50,000 to run while averting $65,000 in treatment and productivity costs, giving net cost of −$15,000; because caries outcomes also improve relative to no program, the analysis reports $15,000 in net savings.
Mistake to avoid: Reporting implementation expenditure as net cost without subtracting the included averted medical and productivity costs from the intervention total.
Reference: [12]
43. Cost-effectiveness ratios need a comparator and outcome unit
When an intervention is more effective and more costly than its comparator, the cost-effectiveness ratio divides net cost by the change in health outcomes, such as cost per case prevented. Because cost-effectiveness analysis is inherently comparative, the ratio is interpreted only relative to the specified alternative, using consistent outcome definitions; no universal willingness-to-pay threshold is established.
Worked example: Compared with current school screening practice, a hypothetical program has net cost of $24,000 and prevents 80 additional cases of untreated decay, yielding a ratio of $300 per additional case prevented, interpreted only against that current-practice alternative.
Mistake to avoid: Calling a ratio cost effective without naming the comparator, the outcome unit, or the decision context that frames the interpretation.
Reference: [12]
44. Separate capital investment from recurring costs
Program costing separates one-time fixed investment, such as fluoridation facilities, from annual recurrent costs covering maintenance, operation including staff, and monitoring. The Community Guide review describes this split conceptually, and studies reported both capital and recurring figures. A first-year total combines both components, so budgets should identify which portion recurs each year rather than treating the whole amount as continuing expense.
Worked example: A hypothetical fluoridation system's equipment costs $40,000 once, while first-year maintenance, operator salaries, and monitoring cost $12,000; the $52,000 first-year total therefore includes a $12,000 recurring component that future annual budgets must continue to fund.
Mistake to avoid: Treating a one-time equipment purchase as the complete continuing budget while ignoring annual maintenance, operation, staffing, and monitoring costs.
Reference: [10]
45. Benefit estimates depend on included categories
Reviewed fluoridation economic studies built benefit totals from averted health care costs, averted productivity losses, and other losses such as travel. Because studies included different component counts, the review named component count a main cause of variation; check categories before comparing totals.
Worked example: A county appraisal finds one study counting only averted restorative care and another also counting averted work time and travel; the reviewer reports each component separately rather than quoting a single combined benefit figure.
Mistake to avoid: Treating a treatment-cost-only estimate as equivalent to an estimate also including productivity and travel savings.
Reference: [10]
46. Benefit-cost ratios vary with scale and scope
In the historical fluoridation review, reported benefit-cost ratios generally increased with community population size, mainly because per capita costs fell through economies of scale on the cost side. The number of benefit components included also contributed to variation among ratios. These findings reflect past studies rather than guaranteed local returns, and little evidence was found for communities with fewer than 1,000 residents.
Worked example: A planner considering fluoridation for a town of 900 residents does not import a published large-city ratio; she commissions a local assessment because per capita costs run higher in small, scattered populations.
Mistake to avoid: Quoting a celebrated large-community ratio as if it applies unchanged to a small rural water system.
Reference: [10]
47. Distinguish estimated economic results from observed data
Appraising economic evidence means asking whether benefits were modeled from assumed effectiveness rates and whether costs were estimated or actually observed. In the reviewed fluoridation studies, all benefits were estimated from similar effectiveness assumptions, and four of six studies used estimated rather than actual cost data. Recognizing which figures are modeled helps appraisers judge how strongly conclusions depend on those assumptions.
Worked example: A hypothetical model reports benefits between $50,000 and $90,000 depending on assumed effectiveness, against $60,000 in costs; the appraisal notes the favorable conclusion holds under some assumptions but not all.
Mistake to avoid: Presenting modeled savings derived from assumed effectiveness rates as though they were directly observed savings in a community.
Reference: [10]
Policy analysis and advocacy
48. Define a policy problem precisely
Policy problem identification should synthesize data on burden, frequency, severity, and scope, then state the issue specifically enough to point toward actionable solutions. Naming a broad disease such as poor oral health does not reveal which lever policy could pull; a precise formulation follows the evidence rather than presupposing one favored solution in advance.
Worked example: A health department reviewing survey data frames its problem as limited insurance acceptance by local dentists, documented through appointment availability counts, instead of the vaguer label of poor children's oral health.
Mistake to avoid: Choosing a preferred intervention first and then describing the problem in terms that justify that choice.
Reference: [11]
49. Include alternatives and the status quo
Policy analysis is stronger when it deliberately gathers evidence on competing approaches, opposing viewpoints, and the consequences of keeping current arrangements in place. Reviewing literature, best practices, and what other jurisdictions have done helps, but each option's local applicability must be assessed. Including the status quo as a genuine option clarifies what a proposed change would actually add or remove.
Worked example: For a rural children's dental access problem, an analyst compares a mobile dental van, school-based sealant expansion, transportation vouchers for existing clinics, and the status quo, weighing each option's evidence against local geography and workforce constraints.
Mistake to avoid: Presenting only the analyst's preferred proposal while omitting rival options and the status quo, which makes the comparison appear more decisive than the evidence supports.
Reference: [11]
50. Compare health impact, feasibility, and economic effects
Prioritizing policy options requires rating each against three interrelated criteria: health impact, feasibility, and economic and budgetary effects. Ratings should be grounded in data, yet they remain partly subjective, so the evidence, data-quality concerns, and reasoning behind each rating should be documented. The final choice depends on how the analyst weights the criteria, and that weighting must be made explicit.
Worked example: Comparing a teledentistry triage program against extended clinic hours, one option shows stronger projected reach but faces staffing shortages and uncertain reimbursement, while the other is easier to deliver; the recommendation explains the tradeoff and why impact was weighted above feasibility.
Mistake to avoid: Reporting a policy's low, medium, or high rating as an objective fact without documenting the evidence, data-quality concerns, and reasoning behind the judgment.
Reference: [11]
51. Adoption strategies need operational and jurisdictional detail
Once a policy is prioritized, analysis must move beyond recommending it. The strategy should clarify how the policy will operate, the concrete steps needed for implementation, the stakeholders to inform, and jurisdiction-specific considerations and information needs. A favorable analysis does not itself confer legal authority or establish permission for related activities, so applicable institutional restrictions must be respected.
Worked example: A proposed school sealant policy names the county health department as implementer, outlines staff training, equipment procurement, and consent procedures, and identifies the data school districts need; the analyst notes that a positive recommendation alone does not authorize implementation.
Mistake to avoid: Treating evidence that a policy improves oral health as proof that implementation is already authorized, skipping operational planning and jurisdictional review.
Reference: [11]
52. Evidence gaps can guide research or incremental policy
When a policy is not ready for adoption because evidence on impact or feasibility is insufficient, the analysis should say so honestly. Options include developing a research agenda identifying the key unanswered questions, or pursuing a more incremental policy. Either path still requires evidence, feasibility analysis, and appropriate authorization; uncertainty should be retained rather than converted into confident endorsement or dismissal.
Worked example: Because the per-child cost of a mobile dental van in one county is unknown, the analyst proposes a research agenda on unit costs and considers a narrower pilot in two districts, explicitly retaining the cost uncertainty in the final recommendation.
Mistake to avoid: Converting missing cost evidence into either a confident endorsement of the program or an automatic conclusion that it provides no benefit.
Reference: [11]
Health communication
53. Lead with a clear action in familiar language
Public-facing health materials work best when the most important information appears first and the intended action is stated concretely, using active voice and familiar words. Necessary technical terms should be defined before use. Simplification must preserve accuracy and the conditions relevant to the audience's decision, so plain wording never replaces needed qualifications.
Worked example: A county oral health program revises its sealant flyer so it opens with 'Call the clinic to book your child's sealant visit,' replacing an abstract statement about facilitating utilization of dental services.
Mistake to avoid: Replacing a usable instruction with an abstract description of program goals, like 'facilitate utilization of dental services' instead of telling families whom to call.
Reference: [16]
54. Readability scores do not measure comprehension
Readability formulas estimate difficulty from syllables per word and words per sentence, but they cannot measure comprehension, predict acceptance, or forecast action. Comprehension often falls below reading level, especially under stress, and formulas ignore layout and design. Use them only alongside pretesting with a sample of the intended audience; no favorable reading-grade result guarantees effective communication.
Worked example: A clinic's fluoride leaflet scores at a fifth-grade level, yet pretesting shows readers cannot identify whom to contact for an appointment. The team revises the contact section and retests with users rather than treating the software's favorable reading level as proof the message works.
Mistake to avoid: Approving a material solely because software reports a favorable reading level, without checking whether users understand the action or can act on it.
Reference: [16]
55. Pretest cultural meaning with the intended audience
Cultural appropriateness comes from engaging members of the target audience early and testing how they interpret messages, images, and terms. Subgroups within a minority group can differ greatly, so what works for one may fail for another. Culturally themed graphics alone do not make an untested message appropriate; pretesting reveals mismatches before distribution.
Worked example: A health department designs a poster around a symbol its staff assumes signals dental care. Focus-group pretesting with local parents shows they read the symbol differently, so the team revises the image and retests before distributing the poster to the community.
Mistake to avoid: Assuming culturally themed graphics make an untested message culturally appropriate, or treating a cultural group as internally uniform without asking preferred terms.
Reference: [16]
56. Translation should preserve meaning and tone
Translators should select audience-appropriate expressions rather than translate literally, because idioms and slang often lose meaning. Independent back-translation by someone other than the original translator helps check whether meaning and tone survived, and field testing with intended audience members confirms interpretation. Qualified translators familiar with the audience support this process but do not replace audience testing.
Worked example: A Spanish version of an appointment reminder translates an English idiom word for word, confusing readers. The team asks a qualified translator to use an expression local families actually use, checks it through independent back-translation, and field tests the revised reminder with intended readers.
Mistake to avoid: Treating word-for-word correspondence as proof of an accurate health message, or translating English idioms literally without checking how intended readers interpret them.
Reference: [16]
Collaboration and sustainable partnerships
57. Shared vision needs explicit roles
Partners who interpret a shared vision differently or leave responsibilities undefined can erode mutual trust. Aligning goals does not mean identical priorities; partners may agree to disagree on some aims. What matters is a consistent understanding of the collective purpose and an explicit assignment of who does what, so expectations are met and collaboration remains productive.
Worked example: A county oral health coalition assumed its dental society member would staff screening events while the society expected coalition staff to recruit volunteers; clarifying these duties in a written role matrix restored cooperation.
Mistake to avoid: Assuming that endorsing a coalition's broad mission statement removes the need to specify which partner performs each operational task.
Reference: [15]
58. Partnerships combine resources for joint activities
A genuine partnership consists of organizations sharing a common focus and pooling resources for joint activities, not simply holding seats on a roster. No single structure fits all partnerships; how groups are named, organized, funded, and run depends on their goals, membership, and circumstances. Reviewers should look for complementary contributions that actually enable the shared work.
Worked example: A health department contributed surveillance data, a dental hygiene program supplied student clinicians under faculty supervision, and a food bank hosted outreach events, together launching a rural sealant initiative none could deliver alone.
Mistake to avoid: Listing organizations as partners on grant applications without documenting what resources or responsibilities each actually contributes to joint activities.
Reference: [15]
59. Evaluate a partnership according to its stage
Evaluation emphasis shifts with partnership maturity. Early-stage assessment focuses on process: partner mix, engagement levels, barriers, governance, and leadership, because organizational effectiveness determines survival. Once established, evaluation can extend to organizational change within partners, reach, membership, outcomes, and efficiency. Measures should match the partnership's stated objectives and current developmental stage rather than applying uniform criteria.
Worked example: A two-year-old coalition reviewed attendance, governance documents, and identified barriers, while a decade-old coalition additionally tracked whether its advocacy campaigns reached targeted communities and produced measurable policy outcomes.
Mistake to avoid: Scoring a coalition formed three months ago as ineffective because population-level oral health outcomes have not yet improved.
Reference: [15]
60. Sustainability requires funding diversity and adaptation
Sustainability planning differs by horizon: short-term efforts keep an initiative viable until objectives are met, while long-term sustainability demands proactive planning, diverse funding sources, retained collaborative leaders, marketing, and adaptability. No formula guarantees financial stability; coalitions should continually reassess activity scope against available funding and revisit their mission as conditions change.
Worked example: Facing expiration of its founding grant, a state coalition surveyed member dues potential, pursued conference revenue, and trimmed lower-priority activities, keeping core surveillance briefings running while leadership transitioned to new co-chairs.
Mistake to avoid: Treating one successful multi-year grant award as evidence that the coalition's long-term financial sustainability has been secured.
Reference: [15]
Sources
- ABDPH Policies and Resources
- ABDPH Qualifying Examination
- ABDPH Certifying Examination
- AAPHD and ABDPH
- CDC Principles of Epidemiology in Public Health Practice, third edition
- CDC NHANES Tutorial
- CDC Program Evaluation Framework, 2024
- ASTDD Best Practice Approach
- ASTDD Best Practice Approach
- CDC and Community Preventive Services Task Force
- CDC Policy Analytical Framework
- CDC POLARIS
- HHS Belmont Report
- CMS Medicaid Dental Care
- ASTDD Best Practice Approach
- CDC Simply Put