England’s National Health Service introduced a payment in March 2024 to encourage dentists to see people who had gone two years without NHS dental care. A study of monthly claims across 42 integrated care boards found a more favorable trend in the pooled local model. A separate test on national totals was inconclusive. Here is how the two comparisons fit together.
- The national averages were 375,882 new patients per month before the incentive and 336,746 afterward. The raw difference was 39,136 fewer per month.
- The pooled board-level model estimated a drop at rollout followed by an improvement in the monthly trend. The national time-series model’s intervals included zero.
- In boards with an improving new-patient trend, overall treatment volume generally held steady or fell. The authors suggest some appointments may have shifted between patient groups.
- Paper date
- 20 November 2026
- Area covered
- 42 integrated care boards in England
- Data
- Monthly claims from March 2023 to March 2025
- Policy date in the model
- March 2024
Original publication: 20 Nov 2026 · The date above refers to this brief.
Monthly raw counts of new NHS dental patients
patients per month, averages across England
Three pieces of an interrupted time series
Fit the earlier trend
Estimate how monthly counts were changing before the policy.
Estimate the step
Check for an immediate shift when the policy began.
Estimate the slope change
Measure how the pace of growth or decline changed afterward.
Start with the claims definition of a new patient
The study counted people with no NHS dental visit in the previous two years as new patients. Someone returning after a long gap could therefore qualify. Records came from practices submitting NHS claims. The measure counts people seen under the NHS. It leaves out unsuccessful booking attempts and visits paid for privately.
Source 1 ↗An average can fall while a trend improves
The raw monthly averages were 375,882 before the premium and 336,746 afterward, a difference of 39,136 fewer per month. Those averages describe the levels across two periods. The pooled board model separately estimated an immediate step of −665 new patients and a change of +117 in the monthly slope. Adding 117 to the earlier slope of −76 gives an estimated post-policy slope of about +41 new patients per month for the pooled board model. This uses rounded point estimates. Uncertainty for the sum would need additional model information.
Source 1 ↗The national and local models use different information
The national model fits a single monthly series of counts across England. Its estimate for the slope change was +4,939, with a 95% confidence interval from −9,453 to +19,331. The pooled model uses monthly records for each board and adjusts for local differences. A model using board records can yield a precise estimate while a model of the national total remains inconclusive.
Source 1 ↗What the local results say about capacity
The paper reports positive trend changes in nine of 42 boards, about 21.4% before correction for testing multiple boards. Dorset retained a statistically clear result after the stated correction. In boards with an improving new-patient trend, the total patients and units of dental activity generally held steady or fell. A shift toward patients with a longer gap could explain this pattern. Appointment histories would be needed to tell who gained or lost access.
Source 1 ↗What a wider evaluation could track
Changes in patient charges and other recovery policies could have affected the same time series. An evaluation of capacity could follow total appointment availability and patients’ waiting times as well as new-patient counts. A study could also compare areas that funded additional capacity with areas that changed only the payment rules.
Source 1 ↗Separate the immediate change, trend and treatment volume
Primary journal article. Checked the eligibility definition, statistical methods, main results table, figure descriptions and study limitations.
01Can monthly counts fall while the estimated trend improves?
- What was observed
- The raw post-policy mean was lower. The pooled model estimated an immediate step down and a subsequent improvement in trajectory.
- Where the conclusion stops
- The period means and the change in slope describe different aspects of the data.
Source 1 · Journal Results and main table, Model 2
02Did more new patients mean more dental care overall?
- What was observed
- Total patients and dental activity were generally flat or declining in boards with an improving new-patient trend.
- Where the conclusion stops
- The authors propose a shift between patient groups. Individual booking and attendance records would allow a direct check.
Source 1 · Journal Results and Discussion
Numbers you can inspect
| Measure | Value & unit | Origin & method |
|---|---|---|
| Integrated care boards | 42 boards | Reported Boards included in the pooled analysis. Source 1 · Journal Methods, Data Sources |
| Raw mean before the premium | 375882 new patients per month | Reported Unadjusted national mean in the pre-policy period. Source 1 · Journal Results |
| Raw mean after the premium | 336746 new patients per month | Reported Unadjusted national mean in the post-policy period. Source 1 · Journal Results |
| Difference in raw means | -39136 new patients per month | Calculated Subtract 375,882 from 336,746. Source 1 · Journal Results |
| Pooled pre-policy slope | -76 new patients per month of trend | Reported Estimated baseline monthly slope of board-level counts. Source 1 · Main table, Model 2, Time trend |
| Pooled level change | -665 new patients | Reported Immediate step estimated from board-level data. Source 1 · Main table, Model 2, NPP |
| Pooled slope change | 117 new patients per month of trend | Reported Change from the pre-policy monthly slope. Source 1 · Main table, Model 2, Change in time trend |
| Approximate post-policy slope | 41 new patients per month of trend | Calculated Add −76 and 117. This uses rounded coefficients and gives the point estimate only. Source 1 · Main table, Model 2, Time trend and Change in time trend |
| National slope change | 4939 new patients per month of trend | Reported Change in the slope of the single national series. Source 1 · Main table, Model 1, Change in time trend |
| Boards with positive trend changes | 9 boards | Reported Count from the subgroup analyses before correction. Source 1 · Journal Results, Figure 2 |
| Share of boards with positive trend changes | 21.4 percent of boards | Calculated 9 divided by 42, multiplied by 100 and rounded. This is the count before multiple-testing correction. Source 1 · Journal Results, Figure 2 |
Compare the actual experiments
These studies answer different questions. Read the unit and endpoint before comparing results.
| Study | Unit & setting | Readout | Interpretation boundary |
|---|---|---|---|
| Dental claims and NHS staff surveys Source 1 · Journal Methods and Statistical Analysis | Local care boards and hospital trusts | Claims for an eligible group and organization-level survey scores | Both studies use earlier data to interpret changes after a policy. They track different outcomes. |
| Dental policies and food benefit changes Source 1 · Journal Methods, Outcomes | Dental claims and household survey responses | Counts of new patients and reports of chronic or recent pain | Eligibility matters for both studies. Check who qualifies for each outcome before comparing results. |
The export includes claims, available numbers, methods and source locations. It contains our reading notes and published summaries; it is not raw participant data or an independent reanalysis.
Evidence update · 21 Nov 2026
First publication. Any calculations in this brief use reported figures. The reading notes identify possible follow-up questions. AI source check; no human editorial or clinical review.
What each dental number measures
| Comparison | Estimate or observation | What it describes |
|---|---|---|
| Raw national counts | 375,882 before and 336,746 after | Monthly averages for the two periods. |
| Pooled local step at rollout | −665 (95% CI −928 to −402) | An immediate level change in the model using board counts. |
| Pooled local slope change | +117 (95% CI +58 to +175) | Change in the monthly slope relative to the earlier slope. |
| National slope change | +4,939 (95% CI −9,453 to +19,331) | Change in slope estimated from the national series. |
| Local treatment volume | Generally stable or falling in areas with an improving new-patient trend | An increase in overall access would need a separate measurement. |
Values come from the journal Results and main table. The national series and pooled board-level series have different units of analysis.
Your questions, answered
Does a new patient have to be visiting a dentist for the first time?
The study used a two-year gap in NHS dental care. That definition can include people with earlier visits.
What does +117 measure?
It is the estimated change in the slope of the board-level model. The rollout step is estimated separately.
How can the raw average be lower when the trend improves?
The averages describe the level across each period. The slope change describes how the trajectory changed. A series can fall at rollout and then climb from that lower level.
What would show that overall dental access expanded?
Total available appointments, completed visits and waiting times for new and returning patients could help answer that broader question.
Limits of this interpretation
- A nationwide policy can coincide with other changes in demand, costs and funding.
- The two-year visit gap defines eligibility for the headline measure. The outcome misses people who tried to book unsuccessfully.
- Testing individual boards raises the chance of false-positive findings.
- The calculations use published point estimates. The raw period means and the fitted slope describe different quantities.
Sources & transparency
- Tan, Niven, Gao et al. (2026) — Financial Incentives and Inequalities in Dental Care Delivery
Primary journal article. Checked the eligibility definition, statistical methods, main results table, figure descriptions and study limitations. · Accessed 21 Nov 2026
DOI: 10.1001/jamanetworkopen.2026.42642
Prepared and source-checked with AI for Press-news Team. The source notes identify the material available during preparation. The article received no human editorial or clinical review. The source list supports the reported results and the calculations shown.
Source check: AI source check — primary research and stated source limits
Clinical review: Not applicable to this educational guide
Suggest a correction