International surveys of patient-reported outcome and experience measures (PROMs and PREMs) have become indispensable tools for monitoring people‑centredness of healthcare systems, promoting cross-country benchmarking and learning. For policymakers, choosing the right source of information to answer a policy question is not always straightforward. Key factors to consider include whether different initiatives overlap or complement each other, how their results compare, and what unique insights each offers for understanding and improving the performance of health systems.
For the Patient Reported Indicator Surveys (PaRIS), OECD Member and Partner countries work together to collect harmonised PROMs and PREMs from primary care users and their primary care practices. PaRIS uses a two‑stage approach, first sampling primary care practices and then patients within those practices. This links patients to the practices where they receive care, allowing outcomes and experiences to be analysed both at the individual level and in relation to how care is organised and delivered.
This paper compares five international patient and/or population health surveys: PaRIS, the European Health Interview Survey (EHIS), the Commonwealth Fund International Health Policy (IHP) Survey, the People’s Voice Survey (PVS) and the Survey of Health, Ageing and Retirement in Europe (SHARE), which collect data on PROMs and/or PREMs in addition to other indicators. These five surveys represent major sources of comparable self-reported data currently available across OECD countries and beyond.
This report presents results from a comparative analysis based on (1) a comparison of survey design, methodology, target population and conceptual orientation; and (2) a comparison of indicators. The indicator comparison comprised three steps: domain mapping, item-level crosswalks and response‑scale harmonisation. First, each question was assigned to an overarching domain. For PaRIS, this followed the PaRIS conceptual framework; for other surveys, questions were mapped to the closest PaRIS domain based on the underlying concept measured. Second, questions within the same domain were paired by topic and assessed for conceptual similarity, wording, reference period and response structure. Third, where items were sufficiently comparable, response categories were harmonised into a binary indicator identifying whether respondents selected one of the two most favourable response options. This “top-three-box” approach was used only when the original scales had the same number of categories and the same number of positive response options. The final set of indicators was restricted to cases where observed differences between surveys for the same country and as-close‑as-possible populations could plausibly be interpreted in light of differences identified in the first stage.
Only a limited number of indicators could be harmonised. Of these, a smaller subset including self‑rated health and experienced quality of care was selected for detailed comparison based on substantial country coverage across surveys. The emerging key findings show that:
International surveys collecting data on PROMs and PREMs offer complementary policy insights. Population‑based surveys such as EHIS and SHARE are designed to monitor health status, behaviours and inequalities across broad populations. Experience‑focussed surveys like the IHP survey and PVS, offer timely information on access, affordability and perceived system performance. PaRIS is designed to generate granular and actionable insights on the healthcare system’s performance on elements that matter for people with chronic conditions in primary care such as on care co‑ordination, outcomes, person-centredness and self‑management.
Design features shape each survey’s policy value and analytical potential. PaRIS is the only survey with a nested, multilevel structure that links patient-reported measures to primary care practice characteristics, enabling analyses of variation and practice‑level improvement levers. SHARE’s longitudinal panel design supports analysis of trajectories over time, while rapid‑cycle surveys such as the IHP survey and PVS prioritise timeliness and responsiveness.
Variation in response rates reflect differences in sampling and fieldwork. EHIS achieves comparatively high participation because it is delivered through national statistical systems with established data‑collection infrastructures. PaRIS maintains relatively strong response rates in most countries by recruiting patients through primary care practices using a two‑stage approach. The IHP survey and PVS tend to have lower response rates. SHARE, a longitudinal panel, benefits from following the same respondents across different waves, but requires more complex, resource‑intensive recruitment and follow‑up.
Only a few domains overlap across surveys. PREMs are included primarily in PaRIS, the IHP survey and PVS while PROMs align more closely across PaRIS, EHIS and SHARE. Only a few indicators overlap across surveys such as experienced quality of care, self-rated health and confidence to self-manage.
The share of people reporting positive outcomes differs substantially across surveys, but absolute levels are not directly comparable. In addition to the methodological differences, these differences also arise because surveys use slightly different question wording and response scales. For example, self‑rated health is measured using “excellent to poor” in PaRIS and “very good to very bad” in EHIS, while experienced quality of care is framed as quality ratings in PaRIS and PVS but as satisfaction levels in the IHP survey.
Self‑rated health, measured as the percentage of respondents reporting positive outcomes, showed up to 30 percentage points (p.p.) differences between PaRIS and other surveys after restricting to comparable populations and standardising results. In Romania, 43% of patients with chronic conditions reported good health in PaRIS compared to 13% in EHIS. In Greece, 69% of patients with chronic conditions reported good health in PaRIS compared to 46% in PVS.
Experienced quality of care, measured as the percentage of respondents reporting good quality of care, also showed differences between PaRIS and other surveys reaching up to around 17 p.p. in crude comparisons and remain substantial after restricting to comparable populations and standardising results. For example, in Canada, around 79% of patients with chronic conditions reported good quality of care in PaRIS compared to 62% in the IHP survey.
Indicators are highly sensitive to differences in the concepts captured by survey questions, even when wording appears similar. For example, PaRIS measures confidence to manage one’s health, while PVS measures the extent to which individuals “feel responsible” for managing their health. Although related, these questions capture different aspects of self‑management, resulting in large differences of 20‑60 p.p. across surveys.
These findings underline that differences in levels across surveys reflect differences in survey design, target population and measurement. Therefore, policymakers should avoid directly comparing absolute values across surveys and instead use each source for the policy question it is best suited to answer. EHIS can be used to identify population groups with poor health status or unmet needs, rapid surveys such as the IHP survey or PVS can be used to monitor public perceptions of access and system performance. PaRIS adds a distinctive and policy relevant lens with granular information by linking patient-reported measures with how primary care is organised and delivered. When used thoughtfully and in combination with other international data sources, these surveys can substantially strengthen evidence‑informed policymaking and support the transition toward more people‑centred, high-performing health systems.