This chapter analyses key results from the OECD Patient-Reported Indicator Survey (PaRIS) on health outcomes and healthcare experiences. It compares results for the 19 participating countries across ten key indicators, highlighting where each country is doing well, and where reform may be needed. Results are summarised in the PaRIS Dashboard, with further details provided in indicator-by-indicator analysis. The chapter also examines some of the essential factors that may explain why patients are more likely to have better (or worse) outcomes and experiences, including linkages across indicators. It finds a positive association between experiences of care and health outcomes. It also shows that countries that spend more on health often have better patient outcomes and experiences, but good results can still be achieved at lower levels of spending.
Does Healthcare Deliver?
Abstract
What the PaRIS data tell us
Health system performance has been typically measured by inputs, processes, and outputs. PaRIS data highlight a crucial change in the way to assess health systems: it focuses on patients’ own perspectives, their health outcomes and their healthcare experiences. These outcomes and experiences vary according to how health systems operate, revealing substantial differences in performance and what drives it.
PaRIS results show a relationship between better experience of care by patients and better health outcomes. This is clear in Australia, Switzerland and the United States, where people consistently report positive care experiences and better health results compared to the OECD PaRIS average. Similar patterns are seen in Canada, Czechia, France and Norway where people report mostly positive experiences with their care, along with relatively good health outcomes. However, this link is not always consistent.
About a tenth of the differences in how people experience healthcare can be attributed to country-level factors. For some indicators, such as experienced quality of care and confidence to self-manage, more than a tenth can be attributed to country differences. This could for example reflect variations in national health policies, and the availability, accessibility and quality of healthcare services, as well as other country-level factors beyond the health system.
Countries that spend more on health are more likely to have better patient outcomes and experiences, but good results can still be achieved at lower levels of spending. Health spending per capita is associated with better outcomes and experiences across the key indicators analysed in this chapter (the “PaRIS10”). At the same time, results show it is possible to have good scores at lower levels of health spending. For example, Belgium, Czechia and Switzerland achieve the highest scores for “Experienced Quality” but at very different levels of health spending per person. More broadly, Czechia and Slovenia are two examples of countries that score well across most of the PaRIS10 indicators, while having relatively low health spending per capita.
PaRIS results highlight an opportunity to improve healthcare performance by tailoring care to a person’s gender, age, and socio‑economic status. For example, people with lower education levels tend to have worse health and report less positive experiences with their care. They are more likely to face poorer health outcomes and are less positive about their treatment compared to those with higher education. Eighty-seven percent of people with lower education reported having chronic conditions, compared to 78% in the higher education group. Women also are more likely to report worse outcomes and experiences, despite tending to live longer than men.
This chapter analyses key results from the OECD Patient-Reported Indicator Survey (PaRIS) on health outcomes and healthcare experiences. It compares results for the 19 participating countries for ten key indicators (the “PaRIS10”). The chapter also examines some of the essential factors that may explain why patients are more likely to have better (or worse) outcomes and experiences, including linkages across indicators.
The rationale underpinning the PaRIS survey is a mandate from OECD Health Ministers to make health systems more people‑centred (OECD, 2017[1]). Health system performance has been typically measured by inputs, processes, and outputs, such as health expenditure, number of hospital beds, number of doctor consultations, or morbidity and mortality rates. PaRIS data highlight a crucial change in the way to assess health system performance: it focuses on patients’ own perspectives, their health outcomes and their healthcare experiences. These outcomes and experiences vary according to how primary care operates, also including its interlinkages with other levels of care, revealing substantial differences in performance.
The fundamental question the PaRIS survey explores is this – to what extent are healthcare systems successful in meeting people’s needs? To answer this question, both patient-reported outcome measures (PROMs) and patient-reported experience measures (PREMs) are used. PROMs and PREMs are standardised validated measures which together show how patients perceive their health and functioning, and the healthcare they receive. PaRIS is a collaborative initiative in which the OECD has partnered with patients, healthcare professionals, policy makers, and academics worldwide.
In terms of scope of analysis, the PaRIS survey concentrates exclusively on primary care users. This allows a nuanced understanding of the experiences and outcomes of people who have firsthand experience with primary care services, an area that has often been poorly understood in many countries to date. A special focus is given to people living with chronic conditions, as these people often have worse health outcomes and make up a large proportion of primary care users. Further salient characteristics of the PaRIS survey are summarised in Box 2.1, with Chapter 7 providing an in-depth discussion of the data and methodology used.
Overall results for the PaRIS10 are then presented in Section 2.2. This is followed by analysis of factors that can explain why patients experience better care or health outcomes in Section 2.3. More specific findings for each of the ten indicators are presented in Section 2.4.
Note that this chapter is intended to provide a high-level overview of the survey results, rather than an in-depth discussion of the policy implications of these results. Such discussion is left to the subsequent chapters (Chapters 3 to 6) which, for specific areas of policy concern, combine more detailed analyses with findings from the literature to explore the policy implications of survey results.
See Chapter 7 for an in-depth discussion of the data and methodology used.
Participating countries
19 countries, of which 17 were OECD members: Australia, Belgium, Canada, Czechia, France, Greece, Iceland, Italy, Luxembourg, the Netherlands, Norway, Portugal, Slovenia, Spain, Switzerland, the United States and Wales (United Kingdom). Romania and Saudia Arabia also participated.
Sampling frame
All primary care users aged 45 years and older, across primary care practices in each participating country. Patients are linked to primary care practices, using a nested sampling approach, so that patient outcomes and experiences can be linked to characteristics of primary care practices.
This age range was chosen because of a particular interest on people living with chronic conditions. These people often have worse health outcomes and make up a large proportion of primary care users.
All people aged 45 who have had at least one contact with their primary care practice in the six months prior to the time of sampling were included.
Common principles for data analysis
Age‑sex standardisation. All results in this report are, unless explicitly stated, age‑sex standardised to adjust for countries different demographic profiles. This helps maximise cross-country comparability of results.
Approach to case‑mix adjustment. Results are stratified between patients with and without chronic conditions. Sensitivity analysis was conducted to compare basic age‑sex adjusted country estimates to more detailed case‑mix adjustment on PROMs and PREMs (see Annex Table 2.A.1). This analysis showed that there were no significant differences in most results, highlighting the robustness of basic age‑sex case‑mix adjustment in the report. This approach is chosen to strike a balance between recognising the importance of case‑mix, whilst avoiding overadjustment that risks masking cross-country performance differences that are amenable to policy.
Comparative intervals. Non-overlapping comparative intervals can be interpreted as statistically significant differences at the 5% significance level. For example, they allow statistical comparisons of a country’s aggregate result with the OECD PaRIS average (which is an unweighted average of the 17 participating OECD members).
Variance components. Reflecting the multi-level design of the PaRIS survey, the extent to which variation in results can be attributed to three levels: country, primary care practice and patient levels.
Some summary descriptive statistics
Results span 107 011 primary care users (aged 45 years and older) and 1 816 primary care practices in 19 countries.
82% of respondents reported having at least one chronic condition, with 52% having multiple chronic conditions (two or more conditions).
18% of respondents were aged 75 or older. Just over half the sample (56%) were women.
33% of respondents had a low education level (lower secondary education or less), while 43% had a high education level (bachelor’s university degree or higher).
Tailored implementation of survey to reflect national circumstances
PaRIS implements standardised patient-reported measures across different healthcare systems, balancing strict uniformity with flexibility. Each country brings unique healthcare structures and contexts and different capacities to conduct a large‑scale, complex patient survey. Therefore, the Secretariat and an international consortium of organisations supporting the methodological aspects of the survey development and implementation (the PaRIS-SUR consortium) worked collaboratively with national experts responsible for the implementation in their respective countries (i.e. National Project Managers) on a case‑by-case basis to determine how best to implement the guidelines, ensuring optimal data collection tailored to the specific circumstances of each country (De Boer et al., 2022[2]).
To enhance comparability, the analysis employs a standardisation mechanism to validate patient samples and ensure that the eligible patient population is adequately represented by age and sex. The robustness of the data was also thoroughly tested, such as through sensitivity analyses with additional case mix correctors. Some differences remain in the samples across countries due to variations in health systems, these are highlighted where relevant and further explained in Chapter 7 (Methods). Over time, the OECD remains committed to refining and harmonising methodologies to further strengthen comparability.
Comparability
The PaRIS sample has been validated to be representative of the eligible patient population as far as possible. This is further detailed in Chapter 7.
In Belgium, the relatively low response rate (<5%) could possibly affect the representativeness of the patient sample. Switzerland included patients in the sample through a “consecutive sampling” approach, enrolling them as they contacted the practice until the required sample size was reached. This method meant that patients who visited their primary care practice more frequently had a higher likelihood of being included in the sample. To assess whether this resulted in a much “sicker” population, a sensitivity analysis with chronic conditions as case‑mix variables was conducted, which did not reveal significant differences.
Canada does not have a national registry of practising primary care practices and therefore, a convenience sample of practices was drawn.
In Italy and the United States, eligibility criteria for patient participation differed from the guidelines. In Italy, the sample only covered three regions. Sampling was limited to patients referred to medical specialists in an ambulatory care setting in Tuscany, Veneto, and the Azienda Unità Sanitaria Locale of Bologna (Emilia Romagna Region). In the United States, the sample is nationally representative for people of 65 years and older but data on people between 45 and 65 years are missing. Due to these deviations in eligibility criteria, results for Italy and the United States have been separated from the rest of the sample and should be interpreted with caution.
See Annex 2.A for supplementary dashboards on full sample, people without chronic conditions, and people aged 65 an older with chronic conditions.
The PaRIS10 dashboard presented in this section summarises how countries compare across the PaRIS10, for patients with chronic conditions. Selection of the PaRIS10 indicators involved close consultation with patient representatives, primary care professionals and academics.
Five outcome indicators focus on essential health metrics. The first three cover the physical, mental, and social domains of health; the next two are more cross-cutting measures of overall well-being and health:
Physical health – ability to carry out everyday physical activities, degree of pain and fatigue.
Mental health – perception of quality of life, mood and ability to think, satisfaction with social activities and relationships, emotional distress.
Social functioning – extent to which a person can carry out their usual social activities and roles.
Well-being – degree to which a person feels positive in terms of their mood, vitality and fulfilment.
General health – overall measure of general health.
Five experience indicators capture core aspects of patients’ interactions with healthcare services. The first three cover different aspects of how people‑centred healthcare is; the next two are broader perceptions on medical care quality and degree of trust in the health system.
Confidence to self-manage – degree of confidence to manage one’s own health and well-being.
Experienced care co‑ordination – extent to which a person experiences a seamless and continuous journey through different healthcare practices and settings.
Person-centred care – extent to which a person’s health needs are managed holistically, ensuring their preferences and needs are central to the care received.
Experienced quality – overall measure of how a person rates the care they have received over the past 12 months.
Trust in healthcare systems – degree to which a person trusts the healthcare system overall.
All ten indicators are based on internationally recognised instruments or guidelines, notably PROMIS, the Patient-Reported Outcomes Measurement Information System (PROMIS, 2017[3]), the WHO‑5 Well-being Index (Heun et al., 2001[4]), the Commonwealth Fund International Health Policy Survey (The Commonwealth Fund, 2017[5]), OECD guidelines on measuring trust (OECD, 2017[6]), and P3CEQ, the Person-Centred Co‑ordinated Care Experience questionnaire (Lloyd et al., 2019[7]). All indicators are designed to be applicable for the general population and for those with specific health conditions.
Table 2.1 provides specifications for each of the ten indicators, including interpretation of values and how they are used in this flagship report. Table 2.2 then presents country-level results of these indicators across the 19 participating countries. Later chapters explore specific indicators in more detail, including sub-components for some of these indicators. In Italy, eligibility criteria for patient participation differed from the guidelines, and the sample only covered three regions. Also, in the United States the implemented eligibility criteria for patients (patients aged 65 and over) differed significantly from the guidelines. Because of these deviations in eligibility criteria, results for these countries are shown separately.
Note: For all ten indicators, the proportion (%) of patients who gave a positive response (outcome or experience) is analysed in this chapter. The thresholds for what constitute a positive response were chosen based on ease of interpretation or other evidence from the literature (e.g. the poor mental well-being cutoff used with the WHO‑5 Well-being Index). For the five indicators based on a composite measure, the average patient score is also analysed. In all cases, a higher score is indicative of better results.
Source: OECD PaRIS patient questionnaire.
Percentage of people reporting positive outcomes or experiences and average scores, for primary care users aged 45+ living with one or more chronic conditions (age‑sex standardised results)
Better than the OECD PaRIS average (statistically higher, at 1+ comparative interval above).
Close to the OECD PaRIS average (not statistically different, within one comparative interval).
Worse than the OECD PaRIS average (statistically lower, 1+ comparative interval below).
Note: For all indicators, percentage values are shown, measuring the percentage of people reporting a positive outcome or experience (for example, patients reporting being in good, very good or excellent general health). For five indicators, the average respondent scale score is also shown. See Chapter 2, Table 2.1 for more information on the cutoffs for positive responses and the scales used. Cells are colour-coded to show if values are statistically different to the OECD PaRIS average. In all cases, higher numbers indicate better performance. In Spain, the trust measure was asked specifically about trust in the primary care professionals within people's primary care practice. This variation might influence trust levels in Spain and limit comparability with other countries that refer to trust in the healthcare system more broadly. * Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. ** the United States sample only includes people aged 65 years or older.
Source: OECD PaRIS 2024 Database.
The following key insights can be drawn from the PaRIS10 dashboard:
In terms of patient outcomes, Switzerland, the United States and Canada scored above the OECD PaRIS average for all five PROMs, while Italy, Portugal Romania, Spain and Wales scored close to or below the OECD PaRIS average for each of these outcome measures.
While most primary care users aged 45 and older are positive about their health, sizable numbers of users are less optimistic. On average, around 3 in 10 people see their physical health and general health as only being fair or poor, and a similar share have poor mental well-being (based on the Well-being indicator). People are more likely to be positive about their mental health and social functioning (respectively 83% and 84% were positive, on average).
For patient experiences, Australia, Belgium, Switzerland and the United States scored consistently above the OECD PaRIS average, while Greece, Iceland and Wales scored close to or below this average for the five PREMs.
Most primary care users – around 9 in 10 people on average – have a positive overall experience of care (experienced quality indicator) and see their healthcare as being person-centred. However, around 4 in 10 people do not feel confident in managing their own health and well-being, did not experience good care co‑ordination, and did not trust the healthcare system overall.
Improving patient experiences is a critical goal in itself. But PaRIS results also show that positive patient experiences are significantly associated at the patient-level with better patient-reported health outcomes, although correlations are not always very high. In general, countries with better patient experiences appear more likely to have higher scores for patient outcomes. In other words, people’s assessment of care quality and trust in the healthcare system is closely linked to how healthy they feel. This association is clear in Switzerland and the United States, where people more often report positive care experiences and better health results compared to the OECD PaRIS average. Similar patterns are seen in Australia, Canada, Czechia, France and Norway where people report mostly positive experiences with their care, along with relatively good health outcomes.
However, the link across indicators is not always consistent. In Spain, for example, while people report mostly positive care experiences, there is still room for improvement in patient-reported health outcomes as compared to the OECD PaRIS average. In Romania, average scores for person-centred care and experienced care co‑ordination were high, but patient-reported outcome measures were comparatively low. Such findings can point policy makers to the area of performance most needing improvement in their country.
Sensitivity analysis using case‑mix adjustment of the PaRIS10 produced consistent and similar findings (see Annex 2.A for further details).
Building on these overall insights, the next section further describes survey results to explore factors that may explain why some countries have better patient-reported outcomes and experiences, while also examining the importance of practice and patient level factors.
2.3.1. Patients who experience better care are also more likely to report better health outcomes
The PaRIS10 dashboard showed a general positive association across patient experiences and outcomes at the country level. Closer inspection of the data confirms that patients experiencing better healthcare are more likely to report better health outcomes, notably in terms of physical health, mental health, and well-being. For example, Figure 2.1 shows that people who found care to be highly person-centred (top quartile, with a score above or equal to 20 out of 24) had better health outcomes than people who found care to be less person-centred (bottom quartile, with a score below 13 out of 24). A closer investigation of these indicators finds that involving patients in care decisions is crucial to enhance confidence in their ability to manage their own health (see Chapter 4). When this happens, patients also report better health outcomes.
Note: Scale score values shown for people receiving highly patient-centred care (PC>=20), low patient-centred care (PC<13) and the OECD PaRIS average – for physical health, mental health and well-being. See Table 2.1 for definitions of scales used for each indicator.
Source: OECD PaRIS 2024 Database.
More broadly, patients who report favourably on one indicator are more likely to report favourably across other PROMs and PREMs (Figure 2.2). This indicates a positive, statistically significant pairwise correlation between all 10 indicators, although the correlations between outcomes and care experiences were rarely high.
The association between experienced co‑ordination and person-centredness is the highest, reflecting the natural overlap between these indicators. Associations between well-being and both mental and physical health are also high. Whilst all correlations are statistically significant, not all associations are strong. Such results suggest that while the PaRIS10 indicators are related to one another, they each measure distinct aspects of health outcomes and care experiences.
Note: Pairwise correlations generated using Spearman method, all are statistically significant (p<0.05).
Source: OECD PaRIS 2024 Database.
2.3.2. More than a tenth of the differences in how people experience healthcare can be attributed to country-level factors
The PaRIS10 dashboard showed that both PROMs and PREMs differ largely between countries. Figure 2.3 shows how much of the variation in the results of the 10 indicators can be attributed to differences between countries, between primary care practices and between patients (based on the intraclass correlation: see Chapter 7 for an explanation of this measure in the context of PaRIS).
While patient-level characteristics (such as age, sex and socio‑economic status) account for most of the variation in results, country-level characteristics still account for a large share of result variation for all five PREMs and some of the PROMs. These results reflect the importance of cross-country differences in the availability, accessibility and quality of healthcare services, but also factors beyond the health system. For example, the impact of the country on how people rate the overall quality of care is over twice as large as the impact of primary care practices (15% compared to 7% of the variation in the experienced quality indicator). Country-level factors also account for a large share (16%) of the variation in results for the indicator confidence to self-manage, and on average more than a tenth of the differences across the five PREMs. Country-level factors were also marked for the PROMs on general health (12%), mental health (8%), and social functioning (6%).
For all ten indicators, country-level factors were more important than practice‑level factors. This implies that the country a patient lives in has a greater impact on how they rate care quality than which healthcare practice they see within that country. Still, results do show relatively large practice‑level differences for the indicator experienced quality (7%). Nevertheless, results suggest that policy change at the national level is more likely to lead to meaningful improvements in patient experiences and outcomes than efforts targeted at individual primary care practices. The following pages explore the impact of some specific country-level, practice‑level and patient-level factors in more detail.
Source: OECD PaRIS 2024 Database.
2.3.3. Countries that spend more on health are more likely to have better patient outcomes and experiences, but good results can still be achieved at lower levels of spending
Comparing data for the PaRIS10 indicators with health spending yields additional insights. These results show that countries spending more on health per capita are likely to have better patient-reported outcomes and experiences on average. For example, Figure 2.4, Panel A shows a positive correlation between health spending per capita (adjusted for purchasing powers) and higher scores on the Well-being indicator. Figure 2.4, Panel B shows a similar pattern for the Experienced Quality indicator, with most countries that spent more (less) than the OECD PaRIS average having higher (lower) scores for this PREM. Correlations between the PaRIS10 indicators and healthcare system resources, such as the number of nurses or doctors, show broadly consistent results.
At the same time, the results show that good outcomes are achievable even with lower levels of health spending. For instance, Czechia, Belgium and Switzerland achieve the highest scores for Experienced Quality of care, despite significantly differing levels of health spending per person. Similarly, Iceland and Slovenia report Well-being scores above the OECD PaRIS average while spending relatively little per capita.
Notably, Czechia and Slovenia perform well across most PaRIS10 indicators despite their relatively low health spending. Later chapters delve into the policies and strategies that contribute to these strong results, demonstrating that substantial increases in health spending are not always necessary to achieve high performance.
Note: Values on both axes were divided by the OECD PaRIS average, meaning that 1 equals the OECD PaRIS average and values above and under 1 respectively for above and under the OECD PaRIS average. For GB-WLS, aggregate UK health spending values are used. * Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. **United States sample only includes people of 65 years and older. AU: Australia, BE: Belgium, CA: Canada, CZ: Czechia, FR: France, GR: Greece, IS: Iceland, LU: Luxembourg, NL: Netherlands, NO: Norway, PT: Portugal, RO: Romania, SA: Saudi Arabia, SI: Slovenia, ES: Spain, CH: Switzerland, GB-WLS: Wales, IT: Italy, US: United States.
Source: OECD Data Explorer, WHO Global Health Observatory, OECD PaRIS 2024 Database.
Multivariate regression analyses that include GDP per capita and control for core patient characteristics confirm the positive correlation with health spending. However, the correlation between countries’ health spending per capita and the PaRIS10 indicators is often moderate (less than 0.5 except for Physical health), particularly for the PREMs (Figure 2.5). That is, whilst health spending is associated with better PaRIS10 indicator scores, it is not a guarantee of better patient outcomes or experiences and, it is possible to perform comparatively well at lower levels of health spending.
Note: An asterisk indicates a statistically significant correlation (p<0.05). In multivariate regressions which controlled for GDP per capita and patient characteristics (age, sex, education, chronic conditions), health spending per capita had a statistically significant coefficient for all the PaRIS10 indicators except “experienced quality”.
Source: OECD Data Explorer, WHO Global Health Observatory, OECD PaRIS 2024 Database.
2.3.4. Countries with higher life expectancy do not have consistently better patient-reported outcomes
In contrast to health spending, there is not a clear relationship between life expectancy and the PaRIS10 indicators. This reflects that while people are living longer, this is not always in good health. Figure 2.6 shows, for example, that several countries known for their high life expectancies, also report relatively low levels of physical and mental health, as reported by patients. More generally, these figures do not show a clear correlation between life expectancy at age 45 and physical health or mental health. Indeed, further analysis found that all correlations between life expectancy (at different ages) and patient-reported outcomes were weak, and none of these correlations were statistically significant.
These results show that mere survival is a blunt metric to evaluate health outcomes. This aligns with research that has found survival or mortality rates lack the granularity to inform policy makers of how well healthcare is working for individuals (Slawomirski, van den Berg and Karmakar-Hore, 2018[8]; Abahussin et al., 2019[9]), and indeed why official statistics also make use of morbidity as well as mortality data.
Note: Values on both axes were divided by the OECD PaRIS average, meaning that 1 equals the OECD PaRIS average and values above and under 1 respectively for above and under the OECD PaRIS average. * Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. **United States sample only includes people of 65 years and older. AU: Australia, BE: Belgium, CA: Canada, CZ: Czechia, FR: France, GR: Greece, IS: Iceland, LU: Luxembourg, NL: Netherlands, NO: Norway, PT: Portugal, RO: Romania, SA: Saudi Arabia, SI: Slovenia, ES: Spain, CH: Switzerland, GB-WLS: Wales, IT: Italy, US: United States.
Source: UN World Population Prospects, OECD PaRIS 2024 Database.
2.3.5. Practice‑level characteristics can have an impact on patient experiences
One of the key features of the design of PaRIS is the potential to link patient experiences with primary care practice‑level characteristics. Exploratory regression analyses showed how some practice characteristics are associated with one or more of the five patient-reported experience measures in this chapter (Table 2.3). Among these practice characteristics, the following were statistically significant:
Patients whose primary care practices scheduled consultations of more than 15 minutes. This is particularly important for patients with two or more chronic conditions, for whom having enough scheduled time with the healthcare professional is essential to address multiple health and care issues (see Chapter 3, for further analysis).
Availability of remote consultation (includes phone, video, and other options). This is further analysed in Chapter 4, showing that video consultations or other remote options are significantly associated with higher person-centred care more than having only telephone consultations.
Perception of how well-prepared the practice is to co‑ordinate care with other health and social care practices for patients with one or multiple chronic conditions, as reported by primary care practices.
These significant characteristics are based on analysis of the full sample. Such results are not always seen in within country-level analysis because of limited variation of certain practice characteristics within some countries.
Note: + denotes positive and significant coefficients in multi-level regression analysis, controlled for age, gender, education, number of chronic conditions, health expenditure per capita, and GDP per capita.
Source: OECD Data Explorer and OECD PaRIS 2024 Database.
2.3.6. Tailoring care to a person’s gender, age and socio‑economic status
Despite the impact country-level factors can have on patient outcomes and experiences, it is patient-level characteristics that explain most of the variation in results. Such results highlight how important it is to adapt healthcare to a person’s gender, age, and socio‑economic status (Table 2.4). Perhaps the most striking result is the gender-health paradox: women tend to live longer than men, yet they report consistently worse health outcomes. Women also report worse healthcare experiences than men, other than the indicator confidence to self-manage where values are almost identical. Chapter 5 describes gender and other disparities in more detail, and how these could be reduced.
There are even larger inequalities between socio‑economic groups. Part of this may be explained by the fact that people with lower education are more likely to have chronic conditions: 87% of primary care users aged 45 or older, compared to 78% of those in the higher education group. Still, PaRIS data show that even after adjusting for the number of chronic conditions, socially disadvantaged groups have worse outcomes (discussed further in Chapter 5). People who completed less education also typically report worse healthcare experiences, other than for the indicator Experienced co‑ordination.
Age differences are less clear-cut. In terms of patient outcomes, while people aged 75 or older are less likely than those aged 45‑54 to report favourably in terms of social functioning and their general health, they are more likely to be positive about their well-being and report being in good mental health. Older people are also more likely to report positive healthcare experiences, other than for the indicator confidence to self-manage.
Note: Values for full sample, OECD PaRIS average, including people without chronic conditions (except for indicators experienced co‑ordination and person-centred care, where no data available for people without chronic conditions). Blue cells denote higher values with statistical significance (p<0.05). See Table 2.1 for definitions of scales and cutoffs used for each indicator.
Source: OECD PaRIS 2024 Database.
2.3.7. People living with chronic conditions are more likely to have worse outcomes but have similar healthcare experiences
Results also show the impact of living with chronic conditions on patient outcomes. Of All primary care users who participated in PaRIS, 82% are living with chronic conditions. Furthermore, 52% of patients having multiple chronic conditions (two or more). Managing these conditions is far more complex and resource‑intensive than dealing with a single illness, putting pressure on health systems and healthcare professionals to deliver high-quality, co‑ordinated care.
People living with chronic conditions generally report worse outcomes than those without chronic conditions, as seen across all five PROMs in all participating countries. Similarly, for the three PREMs measuring primary care experiences, people with chronic conditions are, on average, less likely to report positive experiences compared to those without chronic conditions.
However, the differences between these groups are often minimal, particularly in areas like experienced quality and trust in the healthcare system. In some cases, people with chronic conditions report slightly more positive experiences; for example, in five participating countries, they are more likely to trust their healthcare system.
Chapters 3 and 4 delve deeper into these findings, highlighting how well-designed primary care can improve both outcomes and experiences. For instance, further analysis of PaRIS data show that people with multiple chronic conditions report better care co‑ordination when they have had a medication review – where they discuss all their medications with a healthcare professional.
Figure 2.7 and Figure 2.8 highlight some key findings, with more detailed indicator-by-indicator results comparing people with and without chronic conditions in Section 2.3. Supplementary dashboards in the Annex provide a full breakdown of results by country for people without chronic conditions, as well as for all primary care users.
Note: Results not age‑sex standardised.
Source: OECD PaRIS 2024 Database.
Note: No data available for people without chronic conditions for “experienced co‑ordination” and “person-centredness” PREMs.
Source: OECD PaRIS 2024 Database.
2.4.1. Physical health
The indicator measures physical health based on a person’s response to four questions on physical function, pain and fatigue, using the PROMIS Global Scale. The charts here show the scale score for an “average” person with specified characteristics. These results are further broken down by specific characteristics, such as whether respondents have chronic conditions, and by their primary care practice. This score is a T-score metric in which 42 is the cutoff for being in “good” physical health or better (as compared to “fair” health or worse).
In almost all countries, an average person living with chronic conditions perceived their physical health to be above the good-fair cutoff score of 42 (Figure 2.9). These average scale scores were highest in the United States, Switzerland, Canada, Australia, the Netherlands and Norway where values were all significantly higher than the OECD PaRIS average. Still, the average scale score in all countries was below the very good-good cutoff of 50– likely reflecting that the PaRIS survey samples only those using primary care and an older population group, as compared to the general population.
Despite the average person with chronic conditions typically being in good physical health, 3 out of 10 people living with chronic conditions still rated their physical health unfavourably (below the good-fair cutoff of 42), on average across OECD countries. This rises to 4 in 10 people in Portugal, Spain and Wales, and almost 5 in 10 people in Romania – see the PaRIS10 dashboard (Table 2.2) for further details.
In contrast, for the average person without chronic conditions, their physical health score was above the very good-good cutoff in 13 of 19 countries. In all countries, the average scale score for people living with chronic conditions was significantly lower than for those living without chronic conditions (p<0.05).
People with a low education level reported much worse physical health, reflecting in part the higher prevalence of chronic conditions in this group. Women were also in worse physical health compared to men. People aged 45-54 reported being in better physical health than those aged 75+ (see Section 2.3 and Chapter 5 for further analysis).
Within countries there is generally little variation in patients’ physical health across primary care practices, as compared with other PaRIS10 indicators (Figure 2.3). Still, differences in physical health outcomes across primary care practices amounted to about 10 points (equivalent to one standard deviation) in Australia, Belgium, Czechia, Norway, Saudi Arabia and Wales (Figure 2.10).
Note: PROMIS® Scale v1.2 – Global Health component for physical health. is a T-score metric with a range of 16‑68. Two cutoffs are shown: a good-fair cutoff of 42, and a very good-good’ cutoff of 50. *Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. **United States sample only includes people aged 65 years or older. All within country differences between people with and without chronic conditions are statistically significant (p<0.05).
Source: OECD PaRIS 2024 Database.
Note: Excludes practices with less than ten respondents. Every dark dot represents a primary care practice. PROMIS® Scale v1.2 – Global Health component for physical health is a T-score metric with a range of 16‑68, and a good-fair cutoff of 42. *Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. **Practice‑level data for the United States are not available.
Source: OECD PaRIS 2024 Database.
2.4.2. Mental health
The indicator measures mental health based on a person’s response to four questions on quality of life, emotional distress and social health, using the PROMIS Global Scale. The charts here show the scale score for an “average” person with specified characteristics. These results are further broken down by specific characteristics, such as whether respondents have chronic conditions, and by their primary care practice. This score is a T-score metric in which 40 is the cutoff for being in “good” mental health or better (as compared to “fair” health or worse).
In all countries, an average person living with chronic conditions perceived their mental health to be above the good-fair cutoff score of 40 (Figure 2.11). These average scale scores were highest in Saudia Arabia, the United States, Switzerland and Canada, where values were all significantly higher than the OECD PaRIS average – and above the very good-good cutoff of 48.
Despite most people with chronic conditions typically being in good mental health, about 2 out of 10 people living with chronic conditions still rated their mental health unfavourably (below the good-fair cutoff of 40), on average across OECD countries – see the PaRIS10 dashboard (Table 2.2) for further details.
For people without chronic conditions, the average mental health score was above the very good-good cutoff in most countries (16 of 19). In all countries, the average scale score for people living with chronic conditions was significantly lower than for those living without chronic conditions (p<0.05).
People with a low education level reported much worse mental health, reflecting in part their higher prevalence of chronic conditions. Women were also in worse mental health than men, on average. Differences by age group were negligible, with the average scale score for those aged 75+ the same as those aged 45‑54 (see Section 2.3 and Chapter 5 for further analysis).
Within countries there is generally little variation in patients’ mental health across healthcare practices, as compared with other PaRIS10 indicators (Figure 2.3). The largest differences in mental health outcomes across practices were about 7 points, in Australia, Norway and Wales (Figure 2.12).
Note: PROMIS® Scale v1.2 – Global Health component for mental health score is a T-score metric with a range of 21‑68. Two cutoffs are shown: a good-fair cutoff of 40, and a very good-good’ cutoff of 48. * Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. **United States sample only includes people aged 65 years or older. All within country differences between people with and without chronic conditions are statistically significant (p<0.05).
Source: OECD PaRIS 2024 Database.
Note: Excludes practices with less than ten respondents. Every dark dot represents a primary care practice. PROMIS® Scale v1.2 – Global Health component for mental health is a T-score metric with a range of 21‑68, and a good-fair cutoff of 40, higher values represent better mental health. *Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. **Practice‑level data for the United States are not available.
Source: OECD PaRIS 2024 Database.
2.4.3. Social functioning
The indicator social functioning measures how well people can carry out their usual social activities and roles, using a single item from the PROMIS Global Scale. The charts below show the share of respondents reporting positive outcomes (good, very good or excellent, as compared to fair or poor). These results are further broken down by specific characteristics, such as whether respondents have chronic conditions, and by their primary care practice.
On average across OECD countries, 83% of people with chronic conditions reported positive outcomes for their social functioning, with a cross-country range of 73‑93% (Figure 2.13). Saudi Arabia (93%), France (91%), Switzerland and the United States (90%) had the highest share of people with chronic conditions reporting positive outcomes.
For people without chronic conditions, 94% reported positive outcomes, with at least 9 in 10 rating their social functioning favourably in almost all countries (17 out of 19). In all countries, this share of people reporting favourable outcomes was significantly higher than those living with chronic conditions. Differences between people with and without chronic conditions were in general most marked in Wales, Portugal, Romania, Italy and Spain, countries where relatively few people with chronic conditions reported favourable outcomes.
People with a low education level are less likely to report positive outcomes (79%, compared to 89% for those with a high education level). Age also matters, with people aged 75 years or older less likely to report positive outcomes. In both cases, differences reflect to a large part the higher prevalence of chronic conditions amongst older people and those with a lower education level. Differences by gender were much less profound. These results point to the need for policies that address major risk factors for health and encourage healthy ageing, as well as the importance of the social determinants of health. See Section 2.3 and Chapter 5 for further analysis.
Within countries, there can be large variation in patients’ social functioning across primary care practices, more commonly in countries with comparatively low overall values (Figure 2.14). This is particularly notable in Wales and Portugal. Some primary care practices in these countries had outcomes much higher than the OECD PaRIS average, despite the country averages being the lowest among participating countries. Such variation may reflect health inequalities driven by socio‑economic factors but may also be due to differences in access and quality to associated social care services in the vicinity of primary care practices, with further analysis needed to understand the causes.
Note: PROMIS® Scale v1.2 – Global Health. Answer to the question: “In general, please rate how well you carry out your usual social activities and roles [further specified in questionnaire]”, “good, very good or excellent” versus “fair or poor”. * Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. **United States sample only includes people aged 65 years or older. All within country differences between people with and without chronic conditions are statistically significant (p<0.05).
Source: OECD PaRIS 2024 Database.
Note: Excludes practices with less than ten respondents. Every dark dot represents a primary care practice. PROMIS® Scale v1.2 – Global Health. Answer to the question: “In general, please rate how well you carry out your usual social activities and roles [further specified in questionnaire]”, “good, very good or excellent” versus “fair or poor”. * Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. ** Practice‑level data for the United States are not available.
Source: OECD PaRIS 2024 Database.
2.4.4. Well-being
The indicator well-being measures the degree to which a person feels positive in terms of their mood, vitality and fulfilment, based on a person’s response to five questions that comprise the WHO‑5 Well-being index. The charts here show the scale score for an “average” person with specified characteristics. These results are further broken down by specific characteristics, such as whether respondents have chronic conditions, and by their primary care practice.
This score is based on a raw scale 0‑25 that has been converted to a 0‑100 scale, where 50 is a commonly used cutoff for a person having poor mental well-being.
In all countries, the average score for people living with chronic conditions was above the cutoff for having poor mental well-being, based on their perception of their own well-being (Figure 2.15). These average scale scores were highest in the Netherlands, Switzerland, the United States, Iceland and Norway, whose values were all significantly higher than the OECD PaRIS average.
Still, in only these and four other countries (Canada, Slovenia, Luxembourg and Czechia) did the average person living with chronic conditions feel positive about their well-being more than half of the time (a score of 60 or above). Further, about 3 of 10 people living with chronic conditions reported poor mental well-being (below the cutoff score of 50), rising to around 4 in 10 people in Wales, Portugal and Italy – see the PaRIS10 dashboard (Table 2.2) for further details.
In contrast, most people without chronic conditions felt positive about their well-being more than half of the time (a score of 60 or more), and their average scale score was significantly above those living with chronic conditions (p<0.05), in all countries.
People with a low education level reported worse well-being scores, reflecting in part their higher prevalence of chronic conditions. Women also reported a lower well-being on average than men. People aged 75+ reported higher well-being than those aged 45‑54. These findings by patient characteristics – except for differences between age groups – are consistent with those for the mental health indicator (see Section 2.3 and Chapter 5 for further analysis).
Within countries, there is limited variation in patients’ well-being scores across primary care practices as compared with other indicators (Figure 2.3). Still, differences in outcomes across practices amounted to about 20 points in Spain and Wales, and more than 15 points in Australia, Belgium, France, Norway, Portugal and Saudi Arabia (Figure 2.16).
Note: WHO‑5 well-being index with a range of 0‑100. Two cutoffs are shown: a cutoff of 50, below which suggests poor mental well-being, and 60, representing when a respondent feels positive about their well-being more than half of the time. * Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. **United States sample only includes people aged 65 years or older. All within country differences between people with and without chronic conditions are statistically significant (p<0.05).
Source: OECD PaRIS 2024 Database.
Note: Excludes practices with less than ten respondents. Every dark dot represents a primary care practice. WHO‑5 well-being index with a range of 0‑100. Two cutoffs are shown: a cutoff of 50, below which suggests poor mental well-being. * Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. ** Practice‑level data for the United States are not available.
Source: OECD PaRIS 2024 Database.
2.4.5. General health
This indicator general health provides an overall measure of a person’s health, using a single item from the PROMIS Global Scale. The charts below show the share of respondents reporting positive outcomes (good, very good or excellent, as compared to fair or poor). These results are further broken down by specific characteristics, such as whether respondents have chronic conditions, and by their primary care practice.
On average across OECD countries, 66% of people living with chronic conditions rated their general health positively, but with a wide cross-country range of 39‑93% (Figure 2.17). Saudia Arabia had the highest share of people with chronic conditions reporting positive outcomes (93%), with Canada and the United States also having over 80% of these respondents reporting positive outcomes.
Among people without chronic conditions, 91% reported positive outcomes. In most countries (14 of 19), 9 out of 10 people without chronic conditions rated their general health favourably. In all countries, this share of people reporting favourable outcomes was significantly higher than those living with chronic conditions. Differences between people with and without chronic conditions were in general most marked in countries where relatively few people with chronic conditions reported favourable outcomes.
People with a low education level are less likely to report positive outcomes (60%, compared to 78% for those with a high education level). Age also matters, with people aged 75 years or older less likely to report positive outcomes (60% of respondents, compared to 76% for those aged 45‑54). In both cases, differences reflect to a large part the higher prevalence of chronic conditions amongst older people and those with a lower education level. Differences by gender were much less marked. These results point to the need for policies that address major risk factors for health and encourage healthy ageing, as well as the importance of the social determinants of health. See Section 2.3 and Chapter 5 for further analysis.
Within countries there can be large variation in patients’ general health across primary care practices (Figure 2.18). For example, whilst the country averages for general health in Spain and Wales were markedly lower than the OECD PaRIS average, some primary care practices in these countries had outcomes higher than the OECD PaRIS average. Within-country differences were more than 50 percentage points in Portugal and Spain and were 30 percentage points or more in all surveyed countries, except for Australia, Canada, Italy, Romania and Saudi Arabia.
Note: PROMIS® Scale v1.2 – Global Health. Answer to the question: “In general, would you say your health is …”, “good, very good or excellent” versus “fair or poor”. * Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. **United States sample only includes people aged 65 years or older. All within country differences between people with and without chronic conditions are statistically significant (p<0.05).
Source: OECD PaRIS 2024 Database.
Note: Excludes practices with less than ten respondents. Every dark dot represents a primary care practice. PROMIS® Scale v1.2 – Global Health. Answer to the question: “In general, would you say your health is …”, “good, very good or excellent” versus “fair or poor”. * Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. ** Practice‑level data for the United States are not available.
Source: OECD PaRIS 2024 Database.
2.4.6. Confidence to self-manage
The indicator confidence to self-manage measures how confident respondents are in managing their own health and well-being, based on one question from the P3CEQ data instrument. The charts below show the share of respondents reporting positive experiences (confident or very confident, as opposed to somewhat confident or not confident at all). These results are further broken down by specific characteristics, such as whether respondents have chronic conditions, and by their primary care practice.
On average across OECD countries, only 59% of people living with chronic conditions reported being confident to self-manage (Figure 2.19), but with a very wide cross-country range of 24‑92%. France (92%) had the highest share of people being confident to self-manage for people with chronic conditions, followed by the Netherlands (78%), the United States (74%) and Saudi Arabia (73%).
For people without chronic conditions, 70% reported being confident to self-manage. Within-country differences, though, were only statistically significant in 8 of the 19 countries (France, the Netherlands, Czechia, Australia, Canada, Wales, Norway and Romania).
People with a low education level are less likely to be confident to self-manage (56%, compared to 65% for those with a high education level). Age differences were slightly less marked, with people aged 75 years or older less likely to report confidence to self-manage (57% of respondents, compared to 61% for those aged 45‑54). In both cases, differences could reflect in part the higher prevalence of chronic conditions amongst older people and those with a lower education level. There were no discernible differences by gender. See Section 2.3 and Chapter 5 for further analysis.
Within countries, confidence to self-manage varies markedly across primary care practices for some countries (Figure 2.20), notably in Australia, Canada, Norway and Wales, with differences of more than 25 percentage points. Such variation could reflect differences in the share of people with chronic conditions a primary care practice is covering, as well as differences in the quality of and access to primary care across primary care practices. A more in-depth analysis of this indicator finds that involving patients in care decisions is crucial to enhance confidence in their ability to manage their own health. See Chapter 4 for a more in-depth exploration of this issue.
Note: P3CEQ Questionnaire. Response to question: “How confident are you that you can manage your own health and well-being?”, “confident or very confident” versus “somewhat confident or not confident at all”. * Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. **United States sample only includes people aged 65 years or older. All within country differences between people with and without chronic conditions are statistically significant (p<0.05), except for Saudi Arabia, Switzerland, Spain, Belgium, Portugal, Slovenia, Luxembourg, Greece, Iceland, Italy and the United States.
Source: OECD PaRIS 2024 Database.
Note: Excludes practices with less than ten respondents. Every dark dot represents a primary care practice. P3CEQ Questionnaire. Response to question: “How confident are you that you can manage your own health and well-being?”, “confident or very confident” versus “somewhat confident or not confident at all”. *Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. ** Practice‑level data for the United States are not available.
Source: OECD PaRIS 2024 Database.
2.4.7. Experienced co‑ordination
The indicator experienced co‑ordination is based on the P3CEQ data instrument. The scale includes five questions to measure the extent to which a person experiences a seamless and continuous journey through different healthcare practices and settings. The charts below show the average scale score for people living with chronic conditions in each country, and within country by their primary care practice. This score is based on a scale ranging from 0 to 15. Two cutoffs are shown: a value of 7.5, equivalent to a respondent scoring on average 50% or more across the five questions; and a higher cutoff of 10, equivalent to scoring on average 66.6%. Data are not available for people without chronic conditions.
In 6 of the 19 countries, the average score for people living with chronic conditions was below the cutoff value of 7.5, equivalent to them scoring below 50% across the five questions (Figure 2.21). Furthermore, only in Switzerland and Romania this average score was above 10 (equivalent to scoring on average 66.6%). Indeed, sizeable shares of respondents were not so confident that they had experienced good co‑ordination: on average across OECD countries, around 4 of 10 people living with chronic conditions scored below the cutoff value of 7.5 (less than 50%) – see the PaRIS10 dashboard (Table 2.2) for further details.
People with a low education level reported on average slightly better experiences with co‑ordination, as compared to those with a high education level. This contrasts with all other PaRIS10 indicators, where those with a higher education level were more likely to report a positive result. This could reflect healthcare systems successfully managing co‑ordination for people with low education, though it could also be due to higher expectations from those with higher education levels. Age and gender differences were more marked, with older people and men having higher average scores. See Section 2.3 and Chapter 5 for further analysis.
In several countries there is a wide variation in patients experiencing co‑ordination across practices, notably in Portugal, Saudi Arabia, Switzerland and Wales (Figure 2.22). Such variation points to potential differences in the quality of and access to healthcare of primary care practices, as well as related care in other healthcare settings. A more in-depth analysis of this indicator finds that higher use of care planning and having regular medication reviews offer an important opportunity to improve care co‑ordination. See Chapter 3 and 4 for a more in-depth exploration of this issue.
Note: P3CEQ Questionnaire. Response to five questions measuring care co‑ordination. Scale ranges from 0 to 15. Two cutoffs are shown: a cutoff of 7.5, representing the midpoint of the range of the scale, and a cutoff of 10, equivalent to a positive response on average across questions asked. *Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. **United States sample only includes people aged 65 years or older.
Source: OECD PaRIS 2024 Database.
Note: P3CEQ Questionnaire. Response to five questions measuring care co‑ordination. Scale ranges from 0 to 15, higher scores represent better care co‑ordination. *Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. **Practice‑level data for the United States are not available.
Source: OECD PaRIS 2024 Database.
2.4.8. Person-centred care
The indicator person-centred care is based on the P3CEQ data instrument. The scale includes eight questions to measure the extent to which a person’s health needs are managed holistically, ensuring their preferences and needs are central to the care received. The charts below show the average scale score for people living with chronic conditions in each country, and within country by their primary care practice. This score is based on a scale ranging from 0 to 24. Two cutoffs are shown: a value of 12, equivalent to a respondent scoring on average 50% or more across the eight questions; and a higher cutoff of 16, equivalent to scoring on average 66.6%. Data are not available for people without chronic conditions.
In all countries, people living with chronic conditions scored on average above the cutoff value of 12, equivalent to them scoring at least 50% across the eight questions (Figure 2.23). In most countries (13 of 19), this average scale score was also above the higher cutoff of 16 (equivalent to a positive response on average across the eight questions). Scores were highest in Switzerland and the United States and Australia.
In most countries, around 1 out of 10 people living with chronic conditions did not consider they had received person-centred care (below the cutoff value of 12, or 50%). But this share rises to almost 4 in 10 people in Wales – see the PaRIS10 dashboard (Table 2.2) for further details.
People with a low education level have on average a slightly lower experience score, as compared to those with a high education level. A similar difference was observed by gender, with males having a higher experience score on average, and by age groups, with people aged 75 years or older on average reporting a higher experience score. See Section 2.3 and Chapter 5 for further analysis.
Within countries there can be wide variation in patients’ experiences with person-centred care across practices, with highest differences across primary care practices seen in Portugal and Wales (Figure 2.24). Such variation points to potential differences in the quality of healthcare of primary care practices, and potentially also related care in other healthcare settings. A more in-depth analysis of this indicator points to the key role of continuity of information, and particularly the need to repeat information. See Chapter 4 and 6 for more in-depth explorations of this issue.
Note: P3CEQ Questionnaire. Response to eight questions measuring if care is person-centred. Scale ranges from 0 to 24. Two cutoffs are shown: 12, representing the midpoint of the scale range, and 16, equivalent to a positive response on average across questions asked. * Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. **United States sample only includes people aged 65 years or older.
Source: OECD PaRIS 2024 Database.
Note: P3CEQ Questionnaire. Response to eight questions measuring if care is person-centred. Scale ranges from 0 to 24, higher scores represent better person-centred care. *Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. **Practice‑level data for the United States are not available.
Source: OECD PaRIS 2024 Database.
2.4.9. Experienced quality
The indicator experienced quality provides an overall measure of how a person rates the care they have received over the past 12 months, based on a question adapted from the Commonwealth Fund International Health Policy Survey. The charts below show the share of respondents reporting positive experiences (good, very good or excellent, as opposed to fair, poor or not sure). These results are further broken down by specific characteristics, such as whether respondents have chronic conditions, and by their primary care practice.
On average across OECD countries, 87% of people living with chronic conditions reported positive experiences, but with a wide cross-country range of 69‑97% (Figure 2.25). Switzerland (97%), Czechia (96%) and Belgium (95%) had the highest share of people reporting positive experiences for people with chronic conditions.
Results were very similar for people without chronic conditions (89% reporting positive experiences on average), with no statistical difference between people with and without chronic conditions within all countries.
People with a low education level are slightly less likely to report positive experiences (86%, compared to 88% for those with a high education level). Age differences were more marked, with people aged 75 years or older more likely to report positive experiences (89% of respondents, compared to 83% for those aged 45‑54). Women were slightly less likely than men to report positive experiences. See Section 2.3 and Chapter 5 for further analysis.
Within countries, there is often large variation in patient experiences across primary care practices (Figure 2.26). Indeed, this is the PaRIS10 indicator where practice‑level variation is most prominent (Figure 2.3). Within-country differences across practices are particularly notable in Wales and Portugal, with a more than 50 percentage point range amongst practice estimates. Within-country variation amongst practices was also relatively high in Greece, Spain and Iceland, with a more than 30 percentage point range. In all these countries, patients using their better-performing practices had similar experiences to some of the countries with the highest values overall.
Such variation could point to differences in the quality of healthcare across primary care practices, though further investigation is warranted to understand the potential causes. Indeed, in-depth analysis of the factors behind higher quality healthcare systems offer important insights, including time scheduled for consultation and length of relationship with the healthcare provider (see Chapter 3); as well as if practices manage and invest in their relationship with patients and how they help them navigate the system (see Chapter 4).
Note: Response to question: “When taking all things into consideration in relation to the care you have received, overall, how do you rate the medical care that you have received in the past 12 months from your primary care centre?”, “good, very good or excellent” versus “fair or poor”. *Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. **United States sample only includes people aged 65 years or older. All within country differences between people with and without chronic conditions are not statistically significant (p<0.05).
Source: OECD PaRIS 2024 Database.
Note: Excludes practices with less than ten respondents. Every dark dot represents a primary care practice. Response to question: “When taking all things into consideration in relation to the care you have received, overall, how do you rate the medical care that you have received in the past 12 months from your primary care centre?”, “good, very good or excellent” versus “fair or poor”. * Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. ** Practice‑level data for the United States are not available.
Source: OECD PaRIS 2024 Database.
2.4.10. Trust in healthcare system
The indicator trust in healthcare system measures the degree to which a person trusts their healthcare system overall, based on a question adopting OECD guidelines on measuring trust. The charts below show the share of respondents reporting positive experiences (agree or strongly agree that the healthcare system can be trusted, compared to neither agree nor disagree, disagree, or strongly disagree) with specified characteristics. These results are further broken down by specific characteristics, such as whether respondents have chronic conditions, and by their primary care practice.
On average across OECD countries, 62% of respondents with chronic conditions say that they trust their healthcare system, but with a wide cross-country range of 36‑89% (Figure 2.27). Spain1 (89%) had the largest proportion of people with chronic conditions trusting their healthcare system, followed by Saudi Arabia (74%) and Norway (73%). Results were very similar for people without chronic conditions (64% trusting their healthcare system on average), with no statistical difference between people with and without conditions within all countries. As PaRIS only includes people who are using primary care, it may not include those with the lowest levels of trust who refrain from using healthcare services.
People with a low education level are less likely to trust their healthcare system (59%, compared to 67% for those with a high education level). Age differences were similar, with people aged 75 years or older more likely to report positive experiences (66% of respondents, compared to 59% for those aged 45‑54). Gender disparities were large, with women much less likely than men to report positive experiences (58% versus 68%). This points to a need for greater emphasis on policies that ensure women feel safe and respected in healthcare settings.
Across practices within countries, trust in the healthcare system often vary substantially (Figure 2.28), with a range of 10‑20 percentage points amongst practice estimates in most countries, and around 30 percentage points in Portugal and Wales. More in-depth analysis of trust finds that trust in healthcare systems closely reflects personal experiences, such as whether a person has encountered adverse events, poor care co‑ordination, or problematic waiting times. See Chapter 6 for a further discussion.
Note: Response to question: “How strongly do you agree or disagree that the healthcare system can be trusted?”, “strongly agree, agree” versus “neither agree nor disagree, disagree, strongly disagree”. In Spain, the trust measure was asked specifically about trust in the primary care professionals within people's primary care practice. This variation might influence trust levels in Spain and limit comparability with other countries that refer to trust in the healthcare system more broadly. * Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. **United States sample only includes people aged 65 years or older. All within country differences between people with and without chronic conditions are not statistically significant (p<0.05).
Source: OECD PaRIS 2024 Database.
Note: Excludes practices with less than ten respondents. Every dark dot represents a primary care practice. Response to question: “How strongly do you agree or disagree that the healthcare system can be trusted?”, “strongly agree, agree” versus “neither agree nor disagree, disagree, strongly disagree”. In Spain, the trust measure was asked specifically about trust in the primary care professionals within people's primary care practice. This variation might influence trust levels in Spain and limit comparability with other countries that refer to trust in the healthcare system more broadly. *Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. **Practice‑level data for the United States are not available.
Source: OECD PaRIS 2024 Database.
References
[9] Abahussin, A. et al. (2019), “PROMs for Pain in Adult Cancer Patients: A Systematic Review of Measurement Properties”, Pain Practice, Vol. 19/1, pp. 93-117, https://doi.org/10.1111/PAPR.12711.
[11] Cullati, S. et al. (2020), “Does the single-item self-rated health measure the same thing across different wordings? Construct validity study”, Quality of Life Research, Vol. 29, pp. 2593-2604, https://doi.org/10.1007/s11136-020-02533-2.
[2] De Boer, D. et al. (2022), “Assessing the outcomes and experiences of care from the perspective of people living with chronic conditions, to support countries in developing people-centred policies and practices: study protocol of the International Survey of People Living with Chronic Conditions (PaRIS survey)”, BMJ Open, Vol. 12/9, https://doi.org/10.1136/bmjopen-2022-061424.
[10] Goldstein, H. and M. Healy (1995), “The Graphical Presentation of a Collection of Means”, Journal of the Royal Statistical Society. Series A (Statistics in Society), Vol. 158/1, p. 175, https://doi.org/10.2307/2983411.
[4] Heun, R. et al. (2001), “Validity of the five-item WHO Well-Being Index (WHO-5) in an elderly population”, European Archives of Psychiatry and Clinical Neuroscience, Vol. 251/SUPPL. 2, pp. 27-31, https://doi.org/10.1007/BF03035123/METRICS.
[7] Lloyd, H. et al. (2019), “Validation of the person-centred coordinated care experience questionnaire (P3CEQ)”, International Journal for Quality in Health Care 7, pp. 506-512, https://doi.org/10.1093/intqhc/mzy212.
[1] OECD (2017), Ministerial Statement: The Next Generation of Health Reforms, OECD, Paris, https://www.oecd.org/content/dam/oecd/en/events/2024/1/health-ministerial/ministerial-statement-2017.pdf/_jcr_content/renditions/original./ministerial-statement-2017.pdf.
[6] OECD (2017), OECD Guidelines on Measuring Trust, OECD Publishing, Paris, https://doi.org/10.1787/9789264278219-en.
[3] PROMIS (2017), Global Health: A brief guide to the PROMIS© Global Health instruments, Patient-Reported Outcomes Measurement Information System – Global Health, https://www.healthmeasures.net/administrator/components/com_instruments/uploads/PROMIS%20Global%20Scoring%20Manual.pdf.
[8] Slawomirski, L., M. van den Berg and S. Karmakar-Hore (2018), “Patient-reported Indicator Survey (PaRIS): Aligning Practice and Policy for Better Health Outcomes”, World Medical Journal, Vol. 64/3, https://www.wma.net/wp-content/uploads/2018/10/WMJ_3_2018-1.pdf.
[5] The Commonwealth Fund (2017), 2017 Commonwealth Fund International Health Policy Survey of Older Adults, The Commonwealth Fund, https://www.commonwealthfund.org/publications/surveys/2017/nov/2017-commonwealth-fund-international-health-policy-survey-older.
Results shown in this chapter and throughout the report, unless explicitly stated, are age‑sex standardised without any further case‑mix adjustments. This approach is chosen to strike a balance between recognising the importance of case‑mix, whilst avoiding overadjustment that risks masking cross-country performance differences that are amenable to policy.
To check the robustness of results for the PaRIS10 indicators to alternative model specifications, sensitivity analysis was conducted. The default model used with age‑sex standardisation (Model A) was compared with an alternative model (Model B) that additionally adjusts for the prevalence in PaRIS survey respondents to the most common chronic condition groups, namely: high blood pressure; cardiovascular or heart conditions; diabetes (type 1 or 2); arthritis or ongoing problems with back or joints; breathing conditions (e.g. asthma or COPD); depression, anxiety, or other mental health conditions; and cancer (Annex Table 2.A.1). The assessment includes testing the statistically significant differences (at the 95% level) between country estimates based on the comparative intervals (Goldstein and Healy, 1995[10]).
Results were very close across the model specifications. Results with additional case‑mix adjustment were slightly higher (better outcomes and experiences), but differences were not statistically different at the 95% level from results with only age‑sex standardisation in almost all cases. Countries’ relative position across the ten indicators was also almost always unchanged. The only statistically significant differences between Model A and Model B across the indicators were found for Saudi Arabia’s general health and physical health measures. For general health, Saudi Arabia in Model B increased 3.5 percentage points, with no change in ranking, as compared to Model A. For physical health, Saudi Arabia increased 1.4 T-score metric points in the physical health indicator compared to model A, with an upward shift in four positions.
Further sensitivity analysis was also conducted for the PaRIS indicators that are single items using Likert scales: general health, social functioning, experienced quality, trust in healthcare system, confidence to self-manage. In particular, robustness checks were undertaken to compare results with a positive focus and those including the neutral option (Cullati et al., 2020[11]). Comparisons with the OECD-wide indicator on perceived health (for the population aged 15 years and over) also show strong and significant correlations with the PaRIS indicators on general health, physical health and mental health.
Age and sex standardised plus additional case‑mix adjustment for the seven most reported chronic conditions
Better than the OECD PaRIS average (statistically higher, at 1+ comparative interval above).
Close to the OECD PaRIS average (not statistically different, within one comparative interval).
Worse than the OECD PaRIS average (statistically lower, 1+ comparative interval below).
Note: for five indicators, percentage values are shown, measuring the percentage of people reporting a positive outcome or experience (for example, patients reporting being in good, very good or excellent general health). For five indicators, the average respondent scale score is also shown, based on internationally validated scales. See Table 2.1 for more information on the cutoffs for positive responses and the scales used. Cells are colour-coded to show if values are statistically different to the OECD PaRIS average. In all cases, higher numbers indicate better performance. In Spain, the trust measure was asked specifically about trust in the primary care professionals within people's primary care practice. This variation might influence trust levels in Spain and limit comparability with other countries that refer to trust in the healthcare system more broadly.* Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. **United States sample only includes people aged 65 years or older.
Source: OECD PaRIS 2024 Database.
Percentage of people reporting positive outcomes or experiences and average scores, for all primary care users aged 45+ (age‑sex standardised results)
Better than the OECD PaRIS average (statistically higher, at 1+ comparative interval above).
Close to the OECD PaRIS average (not statistically different, within one comparative interval).
Worse than the OECD PaRIS average (statistically lower, 1+ comparative interval below).
Note: for five indicators, percentage values are shown, measuring the percentage of people reporting a positive outcome or experience (for example, patients reporting being in good, very good or excellent general health). For five indicators, the average respondent scale score is also shown, based on internationally validated scales. See Table 2.1 for more information on the cutoffs for positive responses and the scales used. Cells are colour-coded to show if values are statistically different to the OECD PaRIS average. In all cases, higher numbers indicate better performance. For the indicators “Experienced co‑ordination” and “Person-centred care”, no data available for people without chronic conditions. In Spain, the trust measure was asked specifically about trust in the primary care professionals within people's primary care practice. This variation might influence trust levels in Spain and limit comparability with other countries that refer to trust in the healthcare system more broadly. *Data for Italy refer to patients enrolled in outpatient settings for specialist visits in selected regions. **United States sample only includes people aged 65 years or older.
Source: OECD PaRIS 2024 Database.
Percentage of people reporting positive outcomes or experiences and average scores, for primary care users aged 45+ without chronic conditions (age‑sex standardised results)
Better than the OECD PaRIS average (statistically higher, at 1+ comparative interval above).
Close to the OECD PaRIS average (not statistically different, within one comparative interval).
Worse than the OECD PaRIS average (statistically lower, 1+ comparative interval below).
Note: for all indicators, percentage values are shown, measuring the percentage of people reporting a positive outcome or experience (for example, patients reporting being in good, very good or excellent general health). For five indicators, the average respondent scale score is also shown, based on internationally validated scales. See Table 2.1 for more information on the cutoffs for positive responses and the scales used. Cells are colour-coded to show if values are statistically different to the OECD PaRIS average. In all cases, higher numbers indicate better performance. For the indicators “Experienced co‑ordination” and “Person-centred care”, no data available for people without chronic conditions. In Spain, the trust measure was asked specifically about trust in the primary care professionals within people's primary care practice. This variation might influence trust levels in Spain and limit comparability with other countries that refer to trust in the healthcare system more broadly. *Data for Italy is for patients referred to medical specialists in an ambulatory care setting in Veneto, Tuscany and Emilia Romagna regions, these data are not fully comparable and should be interpreted with caution. **United States sample only includes people aged 65 years or older.
Source: OECD PaRIS 2024 Database.
Percentage of people reporting positive outcomes or experiences and average scores, for primary care users aged 65+ with chronic conditions (age‑sex standardised results)
Better than the OECD PaRIS average (statistically higher, at 1+ comparative interval above).
Close to the OECD PaRIS average (not statistically different, within one comparative interval).
Worse than the OECD PaRIS average (statistically lower, 1+ comparative interval below).
Note: for all indicators, percentage values are shown, measuring the percentage of people reporting a positive outcome or experience (for example, patients reporting being in good, very good or excellent general health). For five indicators, the average respondent scale score is also shown, based on internationally validated scales. See Table 2.1 for more information on the cutoffs for positive responses and the scales used. Cells are colour-coded to show if values are statistically different to the OECD PaRIS average. In all cases, higher numbers indicate better performance. In Spain, the trust measure was asked specifically about trust in the primary care professionals within people's primary care practice. This variation might influence trust levels in Spain and limit comparability with other countries that refer to trust in the healthcare system more broadly. *Data for Italy is for patients referred to medical specialists in an ambulatory care setting in Veneto, Tuscany and Emilia Romagna regions, these data are not fully comparable and should be interpreted with caution. **United States sample only includes people aged 65 years or older.
Source: OECD PaRIS 2024 Database.
← 1. In Spain, the trust measure was asked specifically about trust in the primary care professionals within people's primary care practice. This variation might influence trust levels in Spain and limit comparability with other countries that refer to trust in the healthcare system more broadly.