Did someone put off an appointment to leave it for another patient? It is an appealing explanation for one result from a survey of Europeans during the pandemic. A second result complicates the story. People who scored higher on a scale of helping behaviour also reported more postponements initiated by their health care provider. The survey could establish who respondents said made the decision. It could not establish why they made it.
- The questionnaire was completed in May 2022. The researchers published their analysis on 21 November 2026.
- In the main adjusted model, a one-point increase on a five-point altruism scale was associated with a 10.1 percentage-point higher probability of reporting a self-initiated delay, on average.
- The association was strongest in the youngest age group. A separate analysis also found an association with provider-initiated delays.
- Fieldwork
- 4–24 May 2022
- Geography
- Eight European countries
- Original online sample
- 8,112 adults
- Main analysis
- 6,859 adults after excluding provider-initiated delays
Original publication: 21 Nov 2026 · The date above refers to this brief.
A one-point higher altruism score and the probability of reporting delayed care
percentage-point difference in modelled probability
Follow one respondent through the questionnaire
Recall the previous four months
Did a pandemic-related delay affect a GP visit or elective hospital care?
Record who initiated it
The survey offered one choice to cover the respondent’s delays.
Find the right analysis
The main model excluded provider-initiated delays. The authors included them in a separate comparison.
What counted as postponing care?
Respondents looked back over four months. The main analysis counted someone who said they had personally delayed a general-practitioner visit or elective hospital care by at least a few days because of the pandemic. A delay of days and a delay of months could both count. Someone who needed an appointment but never arranged it might interpret the question differently from someone already holding a booking. That distinction matters if the goal is to count everyone with unmet needs.
Source 1 ↗Read the probability estimate in its original units
People answered nine questions about acts of helping, scored on a scale from one to five. In the authors’ model, a one-point higher score corresponded to a 10.1 percentage-point higher probability of reporting a self-initiated delay, averaged across the analysis sample. Its 95% confidence interval ran from 9.0 to 11.2 points. To find the proportion who cancelled, count respondents who reported a delay and divide by the relevant sample size. The age-specific estimates in the chart come from a separate model that allowed the association to vary by age.
Source 1 ↗The 1,253 people outside the main analysis
Of 8,112 respondents, 1,253 said their provider initiated the postponement. That is 15.4% of the original sample. The researchers left them out of the main comparison and brought them back for an additional analysis. Higher altruism scores were associated with both kinds of delay when each was compared with no delay. If self-sacrifice alone accounted for the result, the second association would be hard to explain. Appointment availability and how people recall a decision are other possibilities that this design cannot disentangle.
Source 1 ↗What would an appointment diary add?
Imagine a clinic offers a later date and a patient agrees. On the questionnaire, a person who reported delays could choose just one initiator across their experiences. A diary could separate who first proposed a change, who agreed, the reason given, and when care eventually happened. Eurostat’s wider unmet-needs measure asks people about needed care they did not receive, with categories such as cost, waiting and distance. It answers a different service question. Its figures cannot be compared directly with a probability difference per point on this survey’s scale.
Source 1 ↗Source 2 ↗Who answered and what the probability difference means
Full BMC Public Health article. Checked the survey dates, exclusion rule, outcome wording, reported marginal effects and the alternative analysis including provider-initiated delays. No appointment records or participant data were available to Press-News.
01What does 10.1 percentage points describe?
- What was observed
- The average marginal effect in the adjusted model for a one-point higher altruism score.
- Where the conclusion stops
- The percentage of people who delayed care is a different statistic.
Source 1 · Main analysis, Model 2 in Results
02Could the survey single out self-sacrifice?
- What was observed
- The authors also found an association with provider-initiated postponement in the full sample.
- Where the conclusion stops
- The respondent selected one initiator. The survey did not document the reason behind each appointment change.
Source 1 · Measures and robustness tests, multinomial analysis
Numbers you can inspect
| Measure | Value & unit | Origin & method |
|---|---|---|
| Original panel | 8112 adults | Reported Respondents across eight countries. Source 1 · Methods, Data |
| Provider-initiated group excluded from the main model | 1253 adults | Reported Respondents classified as provider-initiated. Source 1 · Methods, Analysis |
| Main analytic sample | 6859 adults | Reported Participants retained for the main regression. Source 1 · Methods, Analysis |
| Share of the initial sample excluded | 15.4 percent | Calculated 1,253 ÷ 8,112 × 100, rounded to one decimal. Source 1 · Methods, Analysis |
| Overall marginal effect per scale point | 10.1 percentage points | Reported Model 2, 95% interval 9.0 to 11.2 points. Source 1 · Results, main analysis |
| Marginal effect, ages 18–34 | 16.5 percentage points | Reported Model 3, 95% interval 14.2 to 18.8. Source 1 · Main analysis, Model 3 |
| Marginal effect, ages 35–54 | 11.3 percentage points | Reported Model 3, 95% interval 9.6 to 13.0. Source 1 · Main analysis, Model 3 |
| Marginal effect, ages 55 and over | 5.2 percentage points | Reported Model 3, 95% interval 3.4 to 6.9. Source 1 · Main analysis, Model 3 |
| Difference between youngest and oldest marginal effects | 11.3 percentage points | Calculated 16.5 minus 5.2, using the reported point estimates. Source 1 · Main analysis, Model 3 |
Compare the actual experiments
These studies answer different questions. Read the unit and endpoint before comparing results.
| Study | Unit & setting | Readout | Interpretation boundary |
|---|---|---|---|
| Full-sample analysis of who initiated the delay Source 1 · Robustness tests, full-sample multinomial analysis | Individual survey respondent | Comparison of self- and provider-initiated postponements with no postponement | The survey did not collect a separate decision history for each visit. |
| Eurostat unmet health care needs indicator Source 2 · Eurostat, data sources and definitions | Person reporting a need for examination or treatment | Reported unmet need and reasons, including cost, waiting lists and distance | Different wording and denominator prevent direct comparison with a marginal effect. |
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 · 22 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.
How the association varied by age
| Age group | Average marginal effect per scale point | 95% confidence interval |
|---|---|---|
| 18–34 years | 16.5 percentage points | 14.2 to 18.8 |
| 35–54 years | 11.3 percentage points | 9.6 to 13.0 |
| 55 years and over | 5.2 percentage points | 3.4 to 6.9 |
Age-specific estimates from Model 3, reported in the main results. The estimate combining ages (10.1 percentage points) comes from Model 2.
Your questions, answered
Does the study describe access to care in November 2026?
The publication appeared in November 2026. The survey ran in May 2022 and asked about the previous four months.
Was a missed appointment independently verified?
The outcome was the respondent’s report. The researchers had no matching booking record that could reveal whether or when care was eventually received.
How different were the estimates for younger and older adults?
Subtract the estimate for ages 55 and over (5.2 points) from the estimate for ages 18–34 (16.5). The difference in the two reported marginal effects is 11.3 percentage points. Each slope describes the change in probability associated with a one-point difference in score within that group.
Can we tell that people gave up an appointment for someone else?
A high helping score and a reported decision to delay reveal no reason for a particular cancellation. People with higher scores also reported more provider-initiated delays.
Limits of this interpretation
- The non-random online sample used quotas for age, sex and region. The figures are tied to a specific pandemic period.
- Both the helping score and the delay were self-reported in the same survey. Individual motives and actual appointment histories were unobserved.
- One answer for the initiator could cover multiple appointments. The main model excluded people who selected the provider.
Sources & transparency
- Jahn, Sabat and Schreyögg (2026) — Are the nice ones left behind: a European multi-country survey on the relationship between altruism and forgone care
Full BMC Public Health article. Checked the survey dates, exclusion rule, outcome wording, reported marginal effects and the alternative analysis including provider-initiated delays. No appointment records or participant data were available to Press-News. · Accessed 22 Nov 2026
DOI: 10.1186/s12889-026-30486-7 - Eurostat — Unmet health care needs statistics
Official EU-SILC statistical publication. Checked how the measure of self-reported unmet need and the reasons for missing care differ from this study’s pandemic-delay question. · Accessed 22 Nov 2026
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